Matt Angle:

Welcome back to a new episode of Neurotech Pub. In our season finale, we talked with veteran Neurotech CEOs, Marcus Gerhart, Frank Fischer, and Konstantinos Alataris. This was by far our most listened to episode. We wanted to kick off this next season with some more neuro entrepreneurship—this time talking to Carolina Aguilar, the CEO and founder of INBRAIN Neuroelectronics; Brian Pepin, the CEO and founder of Rune Labs; and Kunal Ghosh, the CEO and founder of Inscopix.

These are three very different companies with different business models, but you’ll see that we have quite a few common interests: the intersection of neurotech with traditional medtech and pharma; engaging and educating potential strategics; and navigating a venture world that is interested in, but still unfamiliar with neurotech. At the time of taping, INBRAIN and Paradromics had just raised money.

Since the taping, Rune Labs also announced a new raise of $23 million Series A. That’s $68 million in independent funding for three early stage companies–something that would’ve been unheard of just a few years ago. The funding environment is finally catching up with the technological one. Check out the show notes for relevant updates and links since the time of the recording. If you’re not already following us on Twitter, that’s where you can find out the latest news in our upcoming season. You can also follow us on Spotify, Apple Podcasts, or wherever you get your podcasts. You can also visit us for more information at paradromics.com/neurotechpub. Here’s to a new season. Cheers.

Matt Angle:

Thanks everyone for joining. Brian, I’m guessing you’re drinking coffee. It’s 11:00 AM for me. And so, I’m supporting the official brewery of Paradromics and all of neurotechnology, Jester King, based in Austin, Texas.

Carolina Aguilar:
Awesome.

Matt Angle:
Kunal, you’re in California as well?

Kunal Ghosh:
Yeah, that’s right. Palo Alto—Bay Area.

Matt Angle:
But Carola you’re in Spain.

Carolina Aguilar:
I’m in Switzerland, actually.

Matt Angle:
Oh, okay.

Carolina Aguilar:
The company is in Spain, but I’ve been living 13 years here now.

Matt Angle:
Oh, nice. So just to get started, could everyone say your name and give a brief introduction?

Carolina Aguilar:
Brian?

Brian Pepin:
I’ll jump in. I’m Brian Pepin and I’m the CEO of Rune Labs. We’re a software and data platform company that works with what I call clinical brain data. So data like electrophysiology and brain imaging that comes from real humans in the context of clinical trials and in the context of routine care. We especially work with data from neuromodulation devices. We have a partnership with Medtronic, for example, we’re working a lot with their Percept, the brain stimulation device and a lot of these newer devices that are, in addition to providing the traditional open loop neuromodulation stimulation, sensing a lot of brain data and helping clinicians and companies and researchers make sense of that and make it really useful for developing and delivering therapies. So, we work mostly in movement disorders right now and we’re doing some cool initial work in MF, I think as well. And then we have a few pilot projects in treatment resistant depression. A therapeutic landscape.

Matt Angle:
Excellent. Thank you.

Carolina Aguilar:
I can go. Name is Carolina Aguilar. Everybody calls me Carola, very historic. And I’m the co-founder and CEO of INBRAIN Neuroelectronics. And we use graphene, which is a new material—bi-dimensional material to decode brain and nerve signals into medical solutions that we have, several interfaces, cortical deep, and also peripheral—and a system, a smart system that is less invasive than what is out there commercially. And we try to make sense out of the signals in all these places to actually create therapies. So that’s what we do since over a year and a half. So it’s very recent.

Matt Angle:
Thank you. Kunal?

Kunal Ghosh:
Hey Matt. I’m Kunal, Kunal Ghosh. I’m the founder and CEO of Inscopix. We build tools for the research community. So, we primarily have been commercializing an imaging system, a platform for imaging large scale neurocircuit connectivity in freely behaving subjects, spanning mice to macaque monkeys. And our platform today is in use in about 500 to 600 labs across the world, enabling, we hope, a lot of good science, close to 200 publications enabled by the Inscopix platform. And we’re super excited about some of the translational applications of the data as they pertain to neuromodulation, BMI, and of course drug development and more traditional drug development.

Matt Angle:
Thanks. I want to get into questions about neurotechnology and the business of neurotechnology, but first, just as a little bit of an opener, the three of you are thought leaders and the expectation of every thought leader is that you’ll be reading a good book and that you’ll share it with our millions of listeners. And so, I’m curious to know, what are you reading right now? Or, in my case, what are you pretending that you have time to read? And maybe tell us a little about that. Or what do you aspire to read?

Brian Pepin:
I’m just finishing up an Ursula K. Leguin book called The Dispossessed, which is really good. Ursula K. Leguin is a fantastic author in the science fiction vein. And…let’s see, it’s a book about an alternative solar system where the moon would be habitable. And then there’s a social revolution on our earth and a bunch of people form a socialist anarchist society on the moon. And then they’re isolated for like 200 years, basically. So the societies develop independently except for they can send letters back and forth. And it’s mostly physicists for some reason, sending letters back and forth. And then the societies come back together through this one guy. And it’s about one physicist basically.

Matt Angle:
Interesting. Thanks.

Carolina Aguilar:
I think that’s going to happen quite soon actually it seems. Everybody is on the moon these days.
I can go. I want to read Sapiens. I haven’t, although when you have kids it’s always a challenge. And I’m revisiting, actually I have it here. There is this one that I just bought. It’s called The Tree of Knowledge. So, this one is before Sapiens because I’ve got kids at home.
And this one I really recommend to everybody. It’s called The Book of Life and I have read it one time, but it is a book that you have to read all your life, I think. That’s why it’s called The Book of Life. Now it’s from Jiddu Krishnamurti, that was an Indian philosopher, and talks about different problems of humanity. And for instance, one of the things about love, love is not retaining–it’s about letting go. And it’s so difficult sometimes when you love something to actually let that person or that thing go. So yes, you’re right. A lot of homework to do, but we are fortunate that others are writing all these amazing things for us.

Kunal Ghosh:
I have to admit from my end that reading, ends up being a luxury and it probably is the same for all of you. But the one book that I’m trying to get to—I’m a bit of a biography junkie—is Barack Obama’s a Promised Land. I really loved his Dreams from My Father and Audacity of Hope and I’m trying my best to get through volume one. So that’s what I’m reading now.

Brian Pepin:
What are you reading, Matt?

Matt Angle:
I’ll tell you what I’m trying to read. It’s called The Open Society and Its Enemies by Karl Popper. So you know, Karl Popper was as a philosopher of science and he wrote at a time where like Freudianism was still being discussed seriously, historicism, and a lot of other pseudoscientific theories were running around at the same time as what we consider science today. And he tried to come up with ways to disentangle what was legit and what wasn’t. And then World War II broke out and he started to look and see that a lot of bad thinking was being hijacked by totalitarian regimes and used to as intellectual props. And so, in startup language, he pivoted his career and started writing about the philosophical and ideological things that prop up totalitarian regimes. And so it’s interesting to me, but I’m not very, far into it.

Carolina Aguilar:
When do you read, do you read at night or early in the morning or weekends?

Brian Pepin:
For me it’s like Saturday and Sunday morning and at night too, but my nighttime reading quality is like is low. Low retention, low comprehension versus the morning reading time.

Kunal Ghosh:
I agree. I think I’m similar to Brian. It’s hard to get time during the weekend and at night. My bedtime reading habits are not that great. So it’s usually on the weekends. Looking forward to some time off during the summer. So hopefully I’ll have more reading time.

Matt Angle:
So another maybe more personal question and now we’re starting to get into some of the venture aspects of our businesses. All of you have venture backed companies and implicitly the venture investors that are putting money into your companies think they’ll be worth hundreds of millions or billions of dollars. And, I’m curious, if you found yourself, some years, hopefully not too many years, from now with a vast sum of money and few responsibilities, what do you think that you would do with it?

Brian Pepin:
Well, obviously, I would be investing in neurotechnology, Matt.

Kunal Ghosh:
Nice one.

Brian Pepin:
The next generation of companies.

Carolina Aguilar:
Give back for sure. There’s so many incredible things to give back. I think investing in neurotech for sure, because when it becomes a passion then it’s there for life. But I have a friend that has spent six months collecting money in Spain and six months investing in Calcutta in India. And I’ve been with her a couple of times and you see that there’s a lot of amazing players that need support. So I guess everybody does a little bit, but then when you have a lot more money, you could do a lot more. So definitely—define a cool project that will support people and see it finished and producing an impact—that would be a dream. Hopefully my kids could also help build it. So that would be one thing.

Kunal Ghosh:
I have to say that my answer is not too much different. Pretty much is exactly what they just said. Giving back would obviously be in the top of the list and investing in innovation and brain science and technologies and projects that can help us obviously continue the journey of helping improve the state of mental health for society. So I think it’s, unfortunately, not too dissimilar. Maybe Matt, you’re looking for a diversity of opinions here, but we’re getting off to a start here where we’re all agreeing with each other.

Matt Angle:
I mean, that was a pretty safe answer. I’m curious though, do you think that you could be investors? I mean, a lot of founders have difficulty becoming investors. It’s in the same way that when you become a manager after being an individual contributor, sometimes it can be difficult to let go of the details. And I think a lot of founders, a lot of executives, sometimes struggle with the transition to being investors where they’re not in charge or they don’t have their hands on the wheel anymore. Do you think you could do that? Because a lot of people I know couldn’t. I’m not sure if I could. I’m not sure if I could get involved investing in a company without wanting to help them. I’m curious if you think that you could easily give someone money and then walk away and say, “Hey, just call me if you need me.”

Kunal Ghosh:
I actually think those are the worst kinds of investors. So I think the best kinds of investors are those that are hopefully prior entrepreneurs or have operating experiences. And maybe the best investors are those that are able to not only invest, but help and drive their investment to a successful outcome by partnering with the company and the people that they’re investing in. And obviously, for of these big multibillion dollar funds where you have an investor sitting on 20 boards maybe they’re not going to be as involved as someone at investing at a seed stage or Series A, or just more founder friendly, more operationally inclined. So I think there are different kinds of investors out there, but for startups and for those going from founding to, I would say mid-stage, I think it’s really important to have investors that are helpful and don’t just give you money and walk away and say, “Call me when you need me.” So I really do think the folks that you’re talking to here have really all the essential ingredients to be good, helpful investors and partners to entrepreneurs.

Carolina Aguilar:
Yes. I think it’s a great question. And I haven’t given much thought except for now, but I tend to trust. So if I would be an investor and I will invest in someone because, at the end, it’s a team, then I have to trust that someone. And I will have done my job in filtering, who do I invest to? But maybe it wouldn’t all be neurotech. I mean, one area of investment that, hopefully, at some point I would like to realize is also the HR models. So, people development models and society models. I think that if we don’t invest in that kind of education, like transformative education, our businesses will be more of what we’ve been doing, like ego-driven and the race for being the best and reinventing the wheel. So maybe that’s the other area? And look, I learn everything about raising money on Wikipedia. So I don’t know, I will have to work a lot to become an investor.

Matt Angle:
I want to talk a little bit about getting VC investment. I’d be curious, when you were first starting out, what percentage of your conversations with venture capitalists led to investment? I’ll lead by saying that it’s probably in the 1-5% range. My experience with a lot of CEOs that I know is that there are a lot of conversations and few of them have translated to investment. But I’m curious to hear, in each of your own journeys, was your first love you married or did you have to meet a lot of investors before you found the right match?

Brian Pepin:
I mean, for me, it’s getting better round by round fortunately. But the first round that we raised, it was probably…I calculated it at some point—it’s less than 2%. I think it was 1.3% or 1.4%. And then the seed round that we raised, we probably seriously talked to probably 20 folks. And then we had one lead and the rest for insiders. So depending on how you look at it, it’s like a 5% hit rate.
I think especially early on, the only way to make money as an angel investor is to be able to invest in deals that are actually good deals, but look like bad deals to everyone else. Because if they’re good deals and they look like good deals, everybody’s going to want in on them and the price is going to be too high and you probably won’t be able to get in on them anyway, unless you have some kind of a special connection. And then obviously if they’re bad, just they’re never going to work out, then you won’t make any money because the company’s going to go to zero. So you’re naturally going to be looking for these deals that look bad to everyone else. Look crazy for whatever reason. Maybe it’s a new team, new entrepreneurs, maybe it’s a new space? Maybe there’s not a predicate market? And I think that just results in being on the other side of that equation as an entrepreneur. A lot of people are going to say “no” to what you’re doing and that’s baked into the system of angel investing.
A lot of people are going to say no, but somebody out there is going to say yes. You’ve got to find that connection. And that’s one of the things that, for me, is so great about being an entrepreneur when you’re trying to pursue a new idea, as opposed to trying to pursue a new idea inside a large company. Like I was at Verily/Google for five years. So, if you have a great new idea for a product or a market that you want to open or whatever, for that company one “no” can kill that. One “no” from a VP or whoever—it takes one “no.” And in angel investing one “yes” is all you need. One person to say “yes.” It’s a much more favorable dynamic for somebody who wants to take risk and do stuff that’s new.

Carolina Aguilar:
In my case, things have happened fast, but I have to say that at the beginning it was poor naiveness. I approached entrepreneurship like a corporate job. There is VCs—let’s visit VCs! But then you realize that there’s a whole plethora of VCs and you have to understand that world. You have to understand there’s the angels, you have to understand that there are medtech [investors], that they are deep tech [investors], and that they are family funds. I mean, you first need to understand because if you just go, you get a lot of “no’s.” But when you start thinking, what is the strategy for funding, you start getting it right. Well, in our case, at the beginning, everybody is like, “No, we need first in human data.” We still get that. But to get to first in human, I need some help now.
But it’s about understanding the ecosystem, understanding the different mechanisms of funding, and going step by step with a very good risk in a strategy. So it is really doing your homework. And I guess at the beginning we just go. And I think it’s good to understand what to do first to prepare and then go. That has been my experience. So just to answer your question, the first round was $1M and it was more like a convertible loan. The second round, it was easier because we already had the investors that knew that if we will de-risk on that first million, they will put more money on the second round, and of course they will help us to help others join. So that’s why we got there $14.3M later.

Kunal Ghosh:
I think, from my perspective, oftentimes what ends up happening, especially entrepreneurs that are first time entrepreneurs as they’re raising, let’s say their seed or pre-seed rounds end up sometimes going to investors that don’t necessarily fit the stage and thesis of the company. And that’s where both parties waste time. And I think it’s really important as you’re starting a company and here, I guess the audience is budding entrepreneurs coming out of universities or companies. And as Brian was saying earlier, obviously clearly great ideas and innovations also do happen in big companies, but it’s easy for those ideas to get killed sometimes. And if there are ideas that make sense to develop outside of a big company, I think leaving a big company and starting a company makes a lot of sense. But those are still, I would say, a little bit more experienced entrepreneurs.
They’ve been in industry, they’ve been through the product journey either—cradle to grave as they say—or at some point of the value chain. And then there are the university-based entrepreneurs, truly the first time entrepreneurs with zero industry experience, zero product development experience and zero raising VC or equity dollars experience. And I’ll count myself as one of those. And when I started the Inscopix journey, about 10 years ago, I had really no product experience. I’d never really had a job. I had been in school all my life prior to then. And I had some ideas about where to raise money and who to go to, but it was my first time starting a company and looking for equity financing.
So I think, looking back, one of the lessons learned is to find the right profile of investors quickly, hone in on the investors that align with your stage and your space. And that saves, I think, both parties a lot of time. So I think for myself, it didn’t end up being a super low hit rate if you will. And I think that’s because I was able to hone in on a group, a profile that would make sense for Inscopix and for the stage we were at. So, I think, for the seed round, we probably talked to 10 groups at most and we were lucky to get investments from two groups and we didn’t raise much, of course. And in our series A we were able to just leverage our network to identify the right profile of investors fairly quickly.
But having said that, I do think that if you look at averages across companies in neurotech, neuroscience, or life sciences, it’s probably exactly what Brian and Carola and you, Matt, have been alluding to—1-5%. But I really do think that ends up being because folks often talk to the wrong investors and it’s probably best to do a little bit of homework first, talk to friends, advisors and just go and focus on the group or groups that truly align with your stage in space. And don’t go talk to the $3B funds, just because they have the name. If you’re raising a million or a $5M round, you probably don’t want to talk to a fund that is investing from their seventh billion dollar fund where the minimum check size has to be $10M for it to make sense for them. So that’s been my experience. Just hone in on the right group, or groups. And I think that enables both parties to be efficient in the fundraising process.

Matt Angle:
Another skill that I think is learned throughout the fundraising process is understanding when you have a productive conversation and when your time is being wasted. I think a complaint that I commonly hear from entrepreneurs is that VCs will waste their time. For instance, Kunal, you are alluding to even when the deal is almost a priori not going to happen, they’ll keep the conversation going to learn more information or maybe do competitive analysis or just write a brief for their partners. I’m curious if you found in your journeys, are there certain tells that someone is wasting your time or that the conversation is not going to go anywhere? Are you getting better at figuring out, “Oh this isn’t serious?” Because I think—

Kunal Ghosh:
Quick answer to this is. At least personally, I haven’t felt that investors have been wasting my time when they’re digging in. And it’s fair as they’re digging in to basically arrive at a conclusion that might not please you as an entrepreneur, but it’s based on them digging in. So I don’t think digging in is necessarily a waste of time, but I do feel that one yellow flag is once an investor is dragging their feet and not coming back to you with either feedback or just a yes/no answer. So I think that to me is sometimes frustrating and that’s where you got to force the investor’s hand, whether or not you like it, whether or not you like what you’ll hear. I do feel that there can be a tendency sometimes for investors to just drag their feet.
And sometimes, honestly, it just could be that schedules are not aligning with their colleagues and people are often on vacation, so there might be genuine reasons sometimes. No question about it. But I do feel, in general, if an investor is taking four weeks to get back to you and is not able to give you quick feedback on a pitch you just gave or questions you just answered based on their diligence then that’s a yellow flag that the interest is probably waning and both parties should probably reach a conclusion pretty quickly.

Carolina:
I didn’t have any experience like that, fortunately. Some might be gathering intel. Yes. I could have sensed that, but in a very proper way, I always had very good feedback about, “Are they interested or not?” Or if we are too early, how to follow up on the deal. And in a very professional and polite way. So far so good.

Matt Angle:
That’s excellent.

Brian Pepin:
I’ve had my time wasted all over the place. All up and down Silicon Valley. I mean, now I’ve had my time wasted enough to where, although it’s a little harder over Zoom, I can tell usually within the first five minutes of talking with somebody if they’re going to be a time waster or not. One of the questions that I’ve started asking—which is a pretty good filter—is what their thesis is around neuroscience or neurotech. And if it’s like, “We don’t have one,” then it’s probably not going to be a good fit for us, to be honest. And so I just try to end that conversation gracefully. Sometimes they have a really interesting thesis and then I can understand how we fit into it. And then that’s a good basis for our conversation. But I don’t know, maybe it’s because we’re more software focused and we’re getting a different attention in the first place? But I’ve had all of the above, people ghosting me, not getting back, saying they’ll get back, saying they’ll make introductions never happening. So, it’s happened to me.

Carolina Aguilar:
Probably it’s more competitive, right? You are in a market that I think is more competitive in neurotech and you have way more business opportunities. So is it a bigger ecosystem? I think in Europe it’s all more concentrated. So therefore probably you have to talk to more people as well.

Brian Pepin:
Yeah, maybe that’s all of it. I hate to throw VC associates under the bus, but if you’re getting contacted by a VC associate, or you get really quickly shoved to an associate, that’s, for me, rarely a good sign. And unless it’s really good, like, sometimes they’ll send you to an associate who has, like, a Neuroscience PhD and you’re like, “Oh, okay, I get this—this is the person who’s done their homework and they’re going to be my champion for the deal process.” But if it’s some random associate, that’s a sign that probably they’re not going to be super engaged.

Kunal Ghosh:
I was actually going to mention that. You’re absolutely right, Brian, I think another yellow flag is when you’re interacting with the associate for weeks and the partner doesn’t have time to do a follow up meeting. I think that is highly correlated with the comment I was making earlier. If you’re finding an investor dragging their feet—they’re asking questions, they’re fair questions, but they’re not either giving you constructive feedback or reaching a decision—I find that that’s highly correlated with the partner not being in the loop and the associate basically saying, “Yep, so and so’s really busy.”
“But here’s another question I have, let’s keep the dialogue going. Here’s yet another question I have.”
And you’re like, “When can we actually have a follow up meeting with, blah, blah, and yourself?”
And, “You know, we’re trying to put something on the calendar.”
So, I think that is, to Brian’s point, a yellow flag when either the group is not closing the conversation in a positive or negative way, nor giving you feedback, just dragging their feet and an associate is carrying the ball all the way. And the partner is not really that involved. I think that’s a yellow flag for sure.

Matt Angle:
I think that’s a piece of advice I wish someone would’ve given me earlier, and that’s something you may not know if you go read Medium pieces about venture capital, a lot of them are written by associates trying to brand themselves. So few people will be like, “Hey, if you’re talking to me, you’re wasting your time.”
But, the truth is that a lot of associates only have the power to say no. They don’t have the ability to say yes or to write a check. They just have the ability to say no. So I want to move on from VC discussions because I want to get into the hearts of your real companies. But if there were one thing that you could go back in time and tell yourself before you started the process, if you just pass one message to old Brian, to old Carola, to old Kunal, what would it be?

Brian Pepin:
I guess mine was just going to be trust your gut, essentially. If this relationship feels bad, it is bad. Move on.

Carolina Aguilar:
Yeah. And to me, it’s practice. But I will also have wished that someone would have trained me a little bit, because I definitely needed a little bit of practice on the first speeches and probably, some of the investors were like, “Go to the billion dollar.” There’s no billion dollar investment here at this stage in Europe I think, but, “Go to the most reputative!” Right. And you go there and if you don’t have the practice of the pitch, of how it works. It’s better to do those at the end, I would say, or later on.

Kunal Ghosh:
You know what I was thinking, and I was intentionally going last to give myself more time to think, I’ll maybe give more generic advice to those that are looking for their first round of financing. And the advice is: try to get customers before you go and actually raise that first fund. I know this is hard for, especially, the companies that are in medtech. So obviously Matt, for a company like Paradromics, that’s tough. I mean, you got to probably work for about four or five years and raise X million dollars before you get to the first milestone of a customer payment. Right? So, for companies that can bring on users early, whether it’s product-based companies or software companies, I think it’s really important to get those first kind of market proof points, and that helped us a lot.
So again, I couldn’t find something to necessarily—there are a lot of things I could have done better, no question about it. But if there were kind of a major piece of feedback that I would give as I reflect upon the journey so far, at least it is helping, having those first customers in place or even having interest if not having the first paying customer in place. I think that gives a lot of credibility to the company. I think that really de-risks the investment for anyone, whether it’s seed or series A, and frankly puts you in a better position to negotiate terms according to where you see your valuation and the company, because you have paying customers and if an investor is not willing to invest, well, okay, thanks, I’m going to keep going because I can bootstrap my way.
And at some point of course, I will need external financing to accelerate progress and to build out the company, but I’m not dependent on external financing to just make payroll because I do have already a revenue source or line of sight to revenue source. Again, I’ll be the first to say that this does not apply to medtech companies where there’s a long journey before that first customer proof point, the first kind of market proof point. But for companies where you do have a market and it’s how do you get to market? I strongly suggest, you know, trying to get to market quickly, even if it’s in a beta phase, get those first users, get those first customers. And that helps a lot with investment. There’s no question about it.

Carolina Aguilar:
Kunal, you said something very, very important apart from what you just said. That is key. What I realized is that also many people don’t take feedback, right? You visit all these customers and they tell you “this sucks”, or “this I love”, or whatever, and many entrepreneurs are getting stuck on their pitch. Right? And if you don’t remove that ego piece and you improve your pitch and you improve your product and you improve all the things that they tell you that might not work, I mean, of course you have to trust in yourself. And I’m not saying that, but feedback and awareness and being able to change your mind and adapt, I think is also key because again, it’s a continuous improvement and all this, that’s why I’m very thankful of all the feedback, especially from the ones that didn’t invest because they made for sure, my pitch and the products better. But many people don’t care and it’s like, yeah, “screw you”, and I continue on my journey. And I think it’s a gift. All those “no”s.

Kunal Ghosh:
I completely agree. I think some of the best feedback that I remember are from those investors that have said no, but have really given very constructive feedback. I often go back and cite their names. I will keep in touch with them. And they’ll be happy that the company is progressing, but their feedback still stays with me as we build a company,

Matt Angle:
Carola, Kunal was just talking about a best-case scenario for a venture-backed business. You’ve got revenue coming in the door and you’d say to you say to potential investors, “Ah, I don’t need you. I’m fine. This is really just growth capital, anyway. I’ll name my terms.” I think you and I are building two very different kinds of businesses. And it’s interesting because I was an idiot. I came from a science background and got into this and then realized I was way over my head. But you’re quite sophisticated. You came from a strong business background. You’d spent time at Medtronic and you decided, “I’m going to build a brain computer interface based on a futuristic material. I’m going to do it.” And so, I would love to know what was it that you saw here? What drove you? What’s your business model? And give us a little bit of your core thesis.

Carolina Aguilar:
It is funny because when this happened, I was in diabetes and it was in and I was in software value based healthcare. So, I was more on the Brian side than on neurotech side. I had gone through that phase and I didn’t see that commercial neurotechnologies were getting sophisticated. So I was losing a little bit of hope. So I said, “okay, I’m going somewhere else”. And at the VC level for other things, I met the founder of INBRAIN. And they asked me, do you want to be CEO of INBRAIN? And I said no. And then I was like, okay, hold on, let’s take your prototypes. And let’s go to see customers, exactly what Kunal is saying. Not as a paying customer, but just to evaluate this thing, right?
So, I look at this, I was like, “Wow! Looks very different to what is out there.” So that caught my attention, but I was not sure that this was going to be the future because we had to work on Biocom, many things, this is all very new, right? But this is when I went to see some of my ex colleagues in Medtronic, the ones that I fully trust and that are like those privileged brains. And I went to see customers in a couple of different countries and they look at this and the data and were like, “Whoa, this changes everything, you have to make this happen.” And I reflect, and I was like, “What the hell? I need to.” I knew that it was hard, but I’ve always lived through that process also in Medtronic.
And I also see that big corporations need help actually to innovate because at the end, in Medtronic, Boston, St. Jude, You have a platform, right? And you have to do incremental steps of innovation at that platform. It is very difficult to start innovating from scratch. So I thought, well, this is what we can do. I actually fill that gap and bring something radically different. And I just could not stop it. I’m an explorer as well. So I’m a crazy explorer and I’m a little bit wild. So I had to take the challenge and yes, that’s what we did.

Matt Angle:
And do you see the med device companies of the world as your potential customers, bringing new technology to them? Or what’s your core proposition?

Carolina Aguilar:
The thing is that now, that is the other angle, right? I’ve always managed business teams and it’s the first time I manage an engineering team, which is, for me, a different planet. But when these people tell you, we discuss deep brain stimulation. We discuss different other technologies that are in the market. And they were like, “This is quite primitive.” And I was like, “Wow!” I thought this was a revolution already when I was working on it for so many years. But you realize that medtech is going to change. It needs to change. So for me, there’s no way back to the “good ol’ medtech.” I think for me, the market is evolving in a completely different direction. And in our case, it’s the intersection between medtech, digital and deep tech, right?
The business models are going to evolve as well. The teams need to evolve. The VC ecosystem is going to evolve. So, when you start combining all these industries is when innovation is happening, but it’s not medtech anymore. There’s no pure systems. I think it’s all getting combined now.

Matt Angle:
And are you proposing to be the next Medtronic or are you proposing to empower Medtronic? Do you see yourself as the vertically integrated next BCI company? Or do you see yourself as having the core technology or what? How are you positioning INBRAIN in the larger VC ecosystem?

Carolina Aguilar:
So, what I learned from my Medtronic experience is that we have to be independent. I can’t build this business to be sold, right? That’s not the purpose. We need to build this to be independent and to compete with whoever is in that market at that given time. And then, if there are good opportunities for integration, for collaboration, for co-development, then we will look at them. But being independent puts all the necessary strategic structures to compete in the marketplace. That is our strategy.

Matt Angle:
And has anything changed in your course strategy since day one? Or have you held true to that core strategy since founding? Have you incorporated feedback? I guess that’s what I’m saying.

Carolina Aguilar:
Yes. Daily. This happens daily. Look, we were with Tom Oxley, right? And I have a huge admiration for Tom. Cause I think what he’s building is also revolutionary, like you guys. We didn’t have time for that. I mean, we are building something that, of course ,is very revolutionary, but it’s also very commercial. I would have wished maybe to have, I don’t know, five years to experiment with graphene in crazy ways. But in our case, and also, because I’m coming from 13 years or 15 years of business experience, it was all about, okay, “What is the first product in the market? And then what is the second?” And there’s less research than in other ventures because the moment you get money from the VCs, you have to put it to work as well. So, I haven’t changed that path, but, of course, I’m incorporating feedback daily on business modeling and potential strategy.

Matt Angle:
Kunal, I’d like to bring you in here for a minute. You started in Inscopix, you were providing miniature microscopes to labs. And one of the things that I’ve been following the Inscopix story for a while. And one of the things that interested me is how service oriented you became, helping people use the microscopes, helping people analyze the data now, even kind of working with strategic partners and pharma companies. Could you tell us a little bit about that journey and how your business vision evolved or, did you have that from the very beginning in mind?

Kunal Ghosh:
Yeah, honestly, a less known fact is we always had that, that to decode the brain for the development of better therapeutics and to influence the design of the next generation neuromodulation, and BMI. And another lesser known fact is, going back to trying to secure customers before you raise your first dollars is our first customer was actually Pfizer. And our first monies in the company was from a big pharma company, honestly, even before the company was even off the ground. And that was, to me, always the arc for the company, for Inscopix, that we got to make drug development in neuro more predictive. There’s no question that there is an ecosystem here that’s being created around neuromodulation and BCI/BMI. And there will be all kinds of uses for these technologies.
But when you look at the burden today of mental illness and brain disease on society, it is enormous. And most of these conditions cannot really be treated well, let alone be cured. And, for better or for worse, today what we have are drug development strategies to go after these indications to really help understand the disease pathophysiology and either directly target a specific cell type, a specific receptor, and/or try to retune a circuit and help really correct the behavioral deficit in the case of mental illness, or help stem the continued progression of disease in the case of neurogenic disease. So, for better or for worse, we’ve got to develop biologics, small molecules. And the pharma-biotech industry is the industry that really needs help.
I like to often quote Tom Insel in this context—he was the former director of the National Institute of Mental Health, and started Mindstrong after spending a couple years at Verily leading their translational neuroscience unit. He often says the house is on fire, and yes, we need to figure out the chemistry of the paint and really revolutionize the understanding and treatment of brain disease, mental illness, but we’ve got to put the fire out. I mean, there are tens of millions of people around the world that are suffering—either diagnosed or undiagnosed—probably hundreds of millions suffering from mental illness and brain disease. And we don’t have great drugs. I mean, just look at the recent debacle with the Biogen drug. And I think some of us know the backstory behind that and how it eventually got approved, but still, approved with a lot of caveats, conditions, and a lot of debate. And that’s the state of drug development today, unfortunately, in Europe.
And when Inscopix entered this back in 2011 and 2012, we wanted to really change the state of drug development by not necessarily getting into drug development ourselves right away, but by realizing that what drug developers from biotech really lacked were good models—good preclinical models that could help the ecosystem of drug developers better understand the disease pathophysiology, screen their compound for efficacy—and when they did get into clinical trials, you know, hopefully they would have a certain confidence that this compound would work or would have such and such risks associated with them. Today, unfortunately, there are really no good predictive models in the neurotherapeutic space. We go into clinical trials pretty much blind, and we find out almost heuristically in phase two or phase three, that the drug will work or not after spending hundreds of millions of dollars and who knows how many years. So, when Inscopix started this, the vision always was to try to not just serve the basic research market, but to really create models, predictive models around Parkinson’s, around Autism; preclinical mouse models, rodent models that could become, in our vision, the gateway from the preclinical phase into phase one.
And when a drug developer in Parkinson’s, for example, would use the Inscopix model and the Inscopix assay, they would go into phase one with, hopefully, a certain degree of confidence that this is not only safe, but also effective in targeting the specific circuit that is implicated in Parkinson’s, in this case the D1-D2 pathways in the basal ganglia, for example. So that was always the raison d’etre for Inscopix. And now of course, when we started in 2011-2012, yes, we had to deliver to Pfizer on the contract we had with them, but we also realized that for this to truly scale and for us to not just have one Pfizer in our journey, but to really empower the community of drug developers, we had to establish the platform as part of mainstream neuroscience.
And that’s how we kind of got into the research market. We realized that, for pharma to really believe in us, we’d first have to get the creme de la creme of neuroscience to believe in us and to appreciate the value of circuit-based neuroscience, to appreciate the, of the data that our platform would generate and publish. And we had that belief that, if we had a hundred labs, and a certain critical mass of science that would be enabled by Inscopix, then there would be credibility and believability that, okay, this platform does have the potential here to be translationally relevant. The data in Parkinson’s, the data in satiety, the data in addiction, the data in, “blah” therapeutic area coming from “blah” lab, or these sets of labs truly have proof points of potentially an assay that could be used in this particular therapeutic area.
So, that’s how we got into the research market. And that’s what drove, as you’re alluding to, Matt, the science-first and service collaboration partner-driven model. That we believed that the value for Inscopix was not in just selling a widget or a product, but in truly connecting the dots between the data and what it would mean for drug development, and, potentially, the development of novel devices for neuromodulation. And that’s what drove us to invest in hiring scientists, in partnering with labs. And it became part of the value proposition, of course.
So, today I think you’re absolutely right that the research community does see us as a service-oriented, science-driven organization. And it’s part of the deal if you will, when they invest in Inscopix, when they buy a product from the company. But the reason for that approach was actually motivated by having our eyes and ears closed to the ground with respect to the science, so that we could start internally figuring out, okay, “There are 10 papers on basal ganglia, and they all seemed to look at these specific circuit patterns, can we now look at all this published data and start to think of how we could build a model for Parkinson’s?” That could be used across the biotech farm ecosystem in the context of Parkinson’s drug development, the same with respect to the hippocampus and memory loss.
So that’s how I think the service oriented mission came about.

Matt Angle:
One of the things that you mentioned, you said, “We wouldn’t work on developing pharmaceuticals right away.” Is that potentially on your roadmap? To do drug discovery out of Inscopix?

Kunal Ghosh:
Potentially. Yeah, I’ll leave it at that. Potentially.

Matt Angle:
I’d be interested because that’s sometimes a classic kind of business challenge, to be developing a platform where you’re bringing in partners, but you might potentially be competing with those partners by utilizing the platform internally. And I’d be curious how you think of that more from a business development standpoint.

Kunal Ghosh:
Hypothetically, what you’re saying is true, but in this space, as you all know, what we’re doing is still so early in the context of the value chain of drug development, that, if we did a better job in identifying a target and potentially even advancing a component to phase one, that would actually be of help to biotech and pharma, rather than them having to do it on their own or with us in a partnership model. So that’s actually what we have found that in neuro, especially pharma is investing less and less in early stage R&D, and is interested in ideally licensing a compound or even buying a company, a biotech that has already proven phase one or phase two efficacy and has met some endpoints. It’s much easier for them to consider drug development and candidates as assets, as opposed to investing in early stage R&D and really going through the journey.
So, we have not found those kinds of competitive dynamics. In fact, we’ve instead found a lot of feedback that has said, “You guys really should try to vertically integrate a little bit more so that you de-risk the journey for us.” And, that’s what we have often found to be more of the case, at least neuro and, and early stage neuro drug development. But you’re right, conceptually speaking, if a platform provider is also starting to get into the application space and competing with its customers that is, you know, potentially a challenge. And there are a lot of companies that navigate this really well, but for us, I don’t think it’s even competitive, to be honest,

Matt Angle:
Brian, you cut your teeth as a hardware engineer and were a senior at Verily and Galvani, and the darling of neural hardware engineering. And then you started a software only company. I’m curious to hear what your thought process was of that.
Brian Pepin:
So, it’s like truism at this point that the future of therapeutics and care delivery is the synthesis of more traditional models with AI, right? That’s what everybody accepts that is true. And I think what I saw at Verily, really from both sides, is that you want one ultimate healthcare product division. It’s going to include, for example, a regulated class three medical device and data of platforms in AI, right? Optimally, the types of companies that it takes to build those things are more or less orthogonal in terms of talent, culture, product development, you name it. they’re different. And so I saw this, like I said, that was on both sides. So inside Verily, DNA of Google, data platforms in AI, very hard to build hardware inside that ecosystem, constantly battling that structure.

Brian Pepin:
And on the other side, we had partners obviously, all around pharma and medtech that were coming to Verily because they didn’t see data platforms as a core competency and had maybe even started something internally that crashed and burned in a big pile of flame and garbage. And come and save us, you know, that’ll help us. And I, for a long time, have been very passionate about human neuroscience in particular
I’ve been very passionate about human neuroscience in particular and any sort of neuroscience therapeutics. And I did my graduate work at Berkeley in that realm of using BMI to investigate Parkinsonian neural circuits and things like that. And, with Galvani, I was really embedded in that world and thought, “Okay, well, there’s these great people that are making next generation hardware. It’s going to take a serious capital investment to make the data platforms and I had to go along with that. Maybe I should go build that type of company after having learned what that looked like at Verily? And then I can invest a lot of effort in it, hire a lot of software engineers, and spread the love across more than just one company so that everybody can benefit from this base level of investment in using the types of really dense electrophysiology and synthesis of imaging and labels that are important to the industry.”
So that was really the kernel of the idea of, “Why go build this company?” And then there were a lot of other decisions here and there. I think another core one was, again, just informed by—going all the way back to grad school, but even informed by my experience at Galvani—that if you’re trying to develop a new therapeutic that has anything to do with the nervous system, forget using animal models, because they are useless. They tell you nothing about efficacy. Maybe they tell you something about safety. Maybe. And so just seeing the barrier, what that felt like internally at Galvani, having those interactions with partners, the pharma-medtech side where there’s like, “Man, we just…we feel like we’re shooting in the dark here until we get into like a phase two equivalent trial.”
And then on the other side, from my neurotech world, seeing again, from the big folks down to the smaller folks, in even into the EEG world, all this great human data that was being created and basically either super siloed, or left on the floor, or stuffed in a Dropbox folder somewhere and forgotten, or not labeled, not aggregated. And that just seemed like that was a core opportunity, both to bring some of the folks that were generating this data more to the center of care. So, if you have a device in epilepsy or Parkinson’s, and you’re generating a bunch of patient specific pathophysiology data, why are you not at the center of care for these patients? That data should be super helpful in caring and everything should feed into this.
That is one, but also, across many, many patients and sites bringing that data back to the research community, both the commercial side and the academic side as a digital laboratory for developing new therapies and developing better biomarkers, better endpoints, better patient phenotyping strategies. Screening targets…maybe after Kunal is finished with them and you have a hundred possible targets or 10 possible targets, what’s the one that you want to try on humans, or the two that you want to try on humans based off that?
And, like I said, you mentioned you were talking with Corola if there’s been pivots. There’s been all these little things along the way. Like I originally expected the concept of a platform to reach a little bit further into the therapeutic, and based on some of the original partnerships and conversations, I think we are now very, very intentionally not doing anything that’s FDA facing. So we’re completely out of the patient risk pathway, and that’s allowed us to develop software a lot quicker, but also engage in the kind of partnerships where there’s this clear line of responsibility where our partners on the hardware side take the patient risk. And we’re the data AI platform that kind of wraps around that.
That was a big decision. Maybe this was somewhat naive. I expected that, just in terms of sales cycles and deal flows, our initial customers were going to be smaller. Our first two customers are Medtronic and Abbott. So that was a bit surprising. We’ve kind of rejiggered the business development process to account for that. I thought clinical trials were going to be the majority of our business for a while. And it turns out that actually working with real world patients, especially in Parkinson’s, is starting to become a much bigger deal. And then we had the global pandemic thrown in there as well, to shake things up. So it’s been somewhat nonlinear, but—

Matt Angle:
And I’m curious–

Brian Pepin:
That’s where we ended up.

Matt Angle:
Where do you see the current demand for the expertise that you’re providing now? Because of course there’s this explosion of new capabilities, new data being generated, but obviously you’re doing a deal with Medtronic and Abbott now. What pain point’s being satisfied right now?

Brian Pepin:
For us, the demand always starts with patients. We’ll just take movement disorders and Parkinson’s. So Parkinson’s patients don’t have disease modifying therapies they can access. There are limitations to current deep brain stimulation therapies. And so there’s a demand for better therapies. And then clinicians want to meet that demand. Clinicians want a better outcome for patients. And so, along comes someone like Medtronic. They have a new DBS device, it’s called the Percept, and it can sense brain signals directly. And there’s promise for using that for adaptive therapies and things like this. Clinicians are excited about it, but it’s new data for them. They’ve never seen it. It’s not really integrated in their clinical practice. And so now, there’s demand for us to come in and provide patients with a new experience.
So we have patients that are using our mobile app and the Apple watch and bringing in a lot of labels along with the brain data, bringing it into a clinical visit where it’s now, the brain data paired with everything else can be used to make decisions by the clinician about how to adjust medications and things like that. And then you can see the effects of those on the other side, on the brain signals of the patients. They can see how local field potentials in the subthalamic nucleus are changing over circadian cycles, before and after medication change. This is before there’s any adaptive neuromodulation component. This is just with the brain single sensing. And so, clinicians like it. And then you get to, “Okay, well if clinicians like it, then they’ll probably prescribe more of these devices.”
And then that’s where there’s a commercial opportunity for us. And then, like I said, there’s that kind of thing. And then there’s a lot of interest for us more on the R&D side as well. How can we bring the types of data that we’re working with? Usually there’s some kind of synthesis going on. So there’s some partner libraries, there’s some platform data that we have, and then there’s some kind of de novo synthesis that’s formed around a biomarker for disease modifying therapy, for Parkinson’s, for example. How can we deploy that in a way where a phase two clinical trial can run with smaller numbers of patients, faster timeframe, higher signal-to-noise ratio, and there’s more likelihood of approval, or faster time to approval, things like that.

Carolina Aguilar:
I have a question for you guys and we don’t have to answer it today. I don’t know if you realize, but if we will put all our companies together, we will have the perfect company. Honestly. And one of the fears that I have, and I’m sure that many of us are doing or have a lot of overlaps on what we are doing, because if at the end we are trying to understand the brain and the code signals and create therapies, even if we have our own areas, there are many, many overlaps.
Do you think, already this generation of people, of leaders, are talking more, collaborating more and doing things different–but talking about the title of the neuropub session today, do you think that we could create new business models that maybe don’t even exist today to leverage all of our strengths and still have a decent business? I mean, how we could compete in different ways from what corporations have been competing at today? Because we cannot keep reinventing the wheel and investing all this money on overlaps. So I know it’s probably a big question and that’s why I said you don’t have to answer it today, but I would love to have time to answer these at some point with people that belong to this generation.

Brian Pepin:
Yeah. I’m coming from a tech ecosystem where there’s integrations everywhere. Take a product like Slack. Slack does its thing, but it integrates with everything under the sun. Asana, Google Docs, all these things. And that makes Slack a better product, but it also makes the customers of these other companies happy. It’s like, “a rising tide lifts all boats.” And I look at that ecosystem and I think there’s a reason that that’s blown so far ahead of other ecosystems around the world, and tech is driving things, is because everybody’s focused on their thing, and as long as the integration points are well defined, it can move forward.
And there’s competition at the edges, but that’s, I think, healthy. And sometimes the integration points shift. I think historically, in medtech, the issue is there hasn’t really been clear integration points. People do these vertical things. And so, for us, we created this integration point where we said, “Okay, well, we’re not going to take any patient risk and that’s going to define that.” But it’s also going to define how we build our product, and where. We have an SDK where folks have a say in what that does, where that sits, and overall product requirements doc for everything. How we sign BAAs with customers, etc.
And for most of our customers, we’re the first person, the first group that’s done that integration at that level. We’re right at that point where things are evolving, and my view is that folks who embrace this model are going to be able to move so much faster that they’re going to win, essentially. Versus the folks who try to do everything totally vertically integrated are just going to be too slow. And, over time, there’ll be that evolutionary process.

Carolina Aguilar:
So, probably this new generation should build more interoperability, right? We should think harder on how to build with each other and not just by ourselves, right? One day you could integrate with graphene, I could integrate with your 1000 channels. We could integrate with Brian, we could integrate with the models. And this is how I think this world will really progress. Right?

Matt Angle:
I think one of the constructs that makes this interoperability in this collaboration difficult, in the med device area, is that regulation is like, “Okay, you tell me exactly what system you’re going to use and exactly what patient, and I’ll approve exactly that system, exactly that patient.” And it gets to be silly. In a modern world, you’d look at some of these approvals and you say like, “well, you have to use this with the computer that’s approved to work with,” and the software and the computer were fit to the device 10 years ago. But I think that there’s going to have to be some structural changes in the way that approvals happen as well, so that we can think about, especially brain-computer interfaces, more broadly.
Brian, I’d be really curious, you kind of sidestep this by saying like, “Oh, we don’t want FDA interaction,” but I think eventually, we’re all going to have to understand that better.

Brian Pepin:
We’re directly working with Class III medical approved devices, and then IDEs that are approved. And, because our software doesn’t affect the patient risk, obviously like HIPAA compliance, GDPR compliance, data privacy, all that stuff, we absorb. And we sign contracts with hospitals and with these companies that manage some of the liability stuff. But, fundamentally, where the language is different from each group, we’re like…founder of the language of an approved supplier or approved vendor. So, there’s some interface that gets set up between us and the partner. Often for us it’s this SDK, software development kit, layer. We define it as much as necessary, but if we’re not changing patient risk, the FDA doesn’t really care as much. So it sits outside that.
There are going to be areas where it makes sense to dip in here or there, and maybe take advantage… At Verily, we were the first group to take advantage of the FDA’s new software as a medical device pre approval. So you approve the whole organization, basically. So you’re essentially approving a development process. And in that context, you can develop software a lot more normally. I think, still though, when we think about partnerships, it’s structurally hard right now to share patient risk, from a regulatory perspective, from a legal perspective. And so I think for the foreseeable future for us, if there is a bucket of patient risk around a device, it’s going to be our partners taking that. And then we have to make sure that we’re pulled out of the loop so we’re not affecting that.
There are some cases like, “Okay, well, what if you take something like an Apple watch and then there’s some label on that that maybe is a certified endpoint or a FDA approved biomarker.” I think there are some things where, okay, well, you might dip in and take some regulatory risks of error, but then that’s somewhat separated from the actual device. It’s a risk, but it’s a different bucket. And there are some areas like that, which are more edge cases that are interesting.
But I agree with you, Matt. If you take something that’s monolithic, like a drug or a Class III medical device, I don’t see a really good way for there not to be ultimately one person who’s ultimately taking, or one company that’s ultimately responsible for that. And then if you’re going to do integrations, either people are going to fall under these suppliers—they’re giving you parts or leads or whatever—or they’re going to be like us, where they’re outside of that risk loop.

Matt Angle:
Brian made a comment earlier about the predictive power of animal models. I would assume to some extent, Kunal, you would take exception with that. And I think for all of us, that’s very relevant because as we talk about developing new markets, it’s really important to understand when you need to move out of an animal model and when you need to get human data. And often the answer is as soon as possible, because some of the biggest market is in mental health, particularly human. I’d be curious to know, what do we see in the state of animal model systems? And, Kunal, do you have hope to offer us?

Kunal Ghosh:
Yeah. I mean, look, we’ve been living in this journey for the last 10 years, right? So firstly, I completely agree with Brian. Animal models have failed us. And this is why neuro is such a, “use your favorite expletive here”. But neuro has unfortunately not progressed anywhere close to the progress we have seen in cardiovascular diseases, oncology, and almost every other therapeutic area. And it’s beyond animal models. At the end of the day, what we lack are quantitative biomarker-driven data sets that can help us, firstly, just to be much more objective about the patient and the progression of disease. Let alone just getting a basic diagnosis. Secondly, biomarkers really start helping with respect to identifying what to target and really helps with the drug development and the device development process. So, for us, this is not…Inscopix was never founded for a quick win. Inscopix was founded to really help transform an ecosystem.
And we’ve already been at this for 10 years and transforming an ecosystem, especially one that is as hard as the neuro ecosystem, neurotherapeutics ecosystem, it’s a long game. It’s by no means a short game. And what we’re fundamentally tackling is literally what Brian said, that animal models have failed us. So how do we make them better? Now, humans are clearly the kinds of subjects and datasets we would want to work with, but how do we open up skulls of hundreds, thousands of humans for the sake of identifying the specific circuits that are going awry? For the sake of identifying the specific cell types that need to be targeted? For the sake of identifying the quantitative biomarkers that will drive not just diagnostics, but also therapeutic development? And that’s why, across the life sciences, not just neuro, we have had to rely on animal models. But, in neuro, why animal models have failed us is because, for the most part, in drug development, the readouts have been crude, low dimensional, behavioral-based readouts.
And there is no question that a mouse that has Parkinson’s and compounds that are correcting the mouse behavior in the Parkinsonian model might or might not work in the context of correcting human behavior. Correcting mouse behavior is by no means indicative of whether human symptomology will be corrected. So I think what we really need in this space is a way to de-risk the clinical trial process by having better animal models that are much more indicative of how the human symptomology would be corrected or would progress. And we fundamentally believe, and we have seen in a lot of these early proof points in some of the data sets that we have generated our customers, colleagues, partners have generated, that the in-brain cell type specific circuit based readout, even in a mouse model that is completely on a different planet in some ways from us as humans, in many diseases that do not rely on neocortex as a readout, as an endpoint, there’s a lot of conservation between mice to humans, as we think about some of the deeper brain structures that are fairly similar across all mammalian species.
Now the neocortex is the one brain region that a mouse will really not come anywhere close to approximating in terms of human circuitry and human disease progression. So, I think there is absolutely truth to what Brian has said, and I frankly don’t take issue with it. I agree with it. And I think what Inscopix is trying to do is turn that old adage around, that animal models suck and have failed us, and try to make them better by not looking at crude, low dimensional readouts that are either behavioral based, but by looking at much more quantitative in-brain circuit based readouts that we tend to think will be conserved.
And if we look at the Rune Labs data, for example, from a Parkinsonian patient, whether it’s a collection of different modalities or just Percept-based readouts. And if we look at similar circuit based data from two or three different mouse models of Parkinson’s, alpha-synuclein, other lesion based models, I wonder if we’re going to see strong correlation in the circuit readouts versus again, relying on just crude, low dimensional behavioral approximations. And that’s the belief. Now, there’s still a long way to go to really test that translational validity. But I can definitely say that in some areas we are already seeing the early signs that the circuit based readout can truly be much more predictive than state of the art, or even other animal based readouts, behavior or non-behavior. And there’s a big question with respect to, can we really get animal models to be much more reliable in terms of predicting efficacy, not just safety? And, frankly speaking, it’s all we have, right?
So there is no question that if we had a simpler way to collect human data, in-brain human data, with specific target identification of drugable targets that we can go after. If we had readouts that we could easily collect and cell types we could easily sequence that we knew were indicative of the disease pathophysiology, I think that’s where we would all be going today. But unfortunately the brain is that one organ that is so hard to access and to find the cell types that are implicated in disease and the RNA targets or the genetic sequences that could be drugable. That’s hard. And I think that’s where we have to come back to preclinical models.
And there still might be ways to bridge between rodent and human with higher order mammalian species, like monkeys. And I think that’s where also our investments in macaque research are aligned to. That, if we can corroborate the same kinds of circuit based data we’re seeing in mice, in macaque monkeys, if we can show literally what Percept is doing in a monkey model, and identify in a close loop manner the kinds of adaptiveness that would be required to truly retune the circuit in a real time manner, in a monkey model, without having to experiment on a human patient. I think that gives device developers a lot of confidence that their protocol, their device, the readouts that they’re seeing in macaque will translate into humans.
So, I think there still has to be a place for preclinical models, but I agree that we have to do a much better job to make them much more predictive, or else we keep flying blind. And I think that unfortunately has been the last 50 years of neuro.

Brian Pepin:
I’ll just say that I am very interested in this potential for like a positive feedback loop between, on the one hand getting finally decent amounts of human data, feeding that back into the animal models, or even the organoids stuff that’s going on, kind of narrowing down, making those better, servicing better possible targets, going back to humans, and doing that loop. Yeah. We haven’t really had the ability to do that before. They’ve been almost two separate worlds. Yeah. A lot of–

Kunal Ghosh:
I think a lot of synergies, exactly, between the novel preclinical platforms and some of the new human data mining, let’s say, platforms. What you guys are doing, what Alto Neuroscience is doing. I think there are a lot of cool synergies in the bridge between human and preclinical, and how each can really complement and influence each other.

Matt Angle:
Can we think of an example of an industry where collaborations have become easy and data sharing has become easy? Because on one hand, I think we agree that integrations could really benefit the neurotech industry. On the other hand, I bet you, each one of us has said at one point during a pitch, “the value’s in the data.” And I think as a result, there’s some hesitancy sometimes for people to share data.

Brian Pepin:
I mean, look at Flatiron, Foundation, Gaurdent. Look at the whole CAR-T therapy industry. You have a mix of folks, you have data platforms, software companies that are handling the complex immunoassay data, genetic data, across thousands of patients, handling those workflows and supply chains. And then you have the pharma companies and biopharma companies, biotech companies, that are developing the regulated stuff. And it’s working pretty well. Mortality rates in skin cancer are down 2% a year over the last 15, 20 years. Largely due to, I think, the adoption of a more data-driven approach. And it’s been the synthesis of new types of companies coming up and making that work. Not to say it wasn’t like an uphill battle at the time.
But, I think part of what changes this a little bit, and the oncology folks have paved away for this too, is that the prevailing, at least in the US, regulatory mindset and political mindset, is that patients own their data. And so if you’re getting consent from the patient at some point, it’s their choice. They can share it with their device company, their pharma company, their hospital. If you can make a value proposition to the patient that you’re going to be a good steward of their data, it’s going to go to something they care about, developing a Parkinson’s therapy, whatever. And then there’s negotiations with the hospitals and clinicians and everything else. But I think taking that kind of patient centric approach–

Carolina Aguilar:
Right? We have all our lives to develop good patient experiences. You know, pharma had all their lives, because I think medtech is more recent, right? But so far, still, after hundreds of years, the patient takes a pill, doesn’t even know which company is providing the pill. And that’s it. And you don’t know anything else from the patient. There’s so many things to improve, and they’re not about sharing data. It is definitely about sharing, but it’s also doing basic things better and putting the patient at the center. That’s the other thing, many people develop things without even talking to the patients or physicians. And I don’t want to…I mean, I love engineers, but in many cases I’ve seen engineers getting amazed by a product, but then they never took to the patient. They never took to the physician. And nobody uses it because nobody has asked, what is that experience with that product, then? What is that experience with that product and does it really work?
So I think there’s a lot of basics to do, and I really believe that collaboration is the new innovation if we want to go fast. And of course data is now going to allow us to do some of those data mining models going faster on a prospective scale in the sense that all these interdependencies are correlations, in silico, animal models, but also human models are going to be key when it comes to data. But yeah, if we will cooperate more on the basics, we will go, I think, double the speed or triple speed.

Matt Angle:
I’ll share a quick anecdote from Frank Fischer, the CEO of NeuroPace. I told him very foolishly that I think that pharma companies are going to start getting plugged in with neural data and interested in neural biomarkers and that there’s going to be a closing gap between pharma and med device. He kind of laughed at me and told me that NeuroPace through all of the data that they’ve collected, they’ve been able to predict when one of the patients with an RNS device is going to respond to the different anti-seizure medications that they’re taking far before you could make a clinical determination of that efficacy. They shared some information with a large pharma company and the pharma company took no interest, took no action.

Frank Fischer:
I’ll share with you about what I know, water magnitude four/five years ago. We were having a conversation with a pharma company about eating around in a restaurant. And this to me is just the reasons. This is one of the few companies that were still interested in developing drugs for epilepsy because that’s common in terms of being in Vogue or out of Vogue and this story at the time, there were only a handful of companies that actually were moving forward with candidates for epilepsy. And one of the things we had shown, by virtue of the data accumulated is that for a patient that has our implant and the changes made by the physician and the medication that patient was taking, we can show in a matter of two weeks whether or not that medication was likely to work with our patient.
And we think about what was the other feedback mechanism for a pharma company. Epilepsy patients are seen every three to six months, physician changes medication, patient comes back three or six months later. And then ask the patient how they are doing and ask the patient previously to keep a manual, seizure file, which frankly, patients are notoriously bad at doing. And that was their feedback mechanism. And I’d say to them, if you’re conducting a drug, I can tell you in phase two whether or not this drug is likely to work. You don’t have to wait until you spend a billion dollars getting to the end of phase three to open that envelope and figure if this stuff works or not. They didn’t do the investment. It’s almost like you’re speaking a foreign language to them because it was so far out of their model that they had been used to. Still amazes me to this day but it is what it is.

Matt Angle:
And I think that that’s largely been par for the course for medical device and pharma companies that they’ve been pretty slow to embrace innovation and the whole healthcare system has moved much slower than the technology field. And I’m wondering, do we think that we’re seeing a fundamental change here? Do you think we can get them to engage and deploy real capital in the field? Or do you think they’re just going to continue putting their toe in the water?

Brian Pepin:
Well, I mean, it’s got to be something they care about, right? I don’t know what Frank’s pitch to these companies was, but it may have just been that, okay if you can detect a seizure earlier, that might be great clinically, but it’s not going to help us develop a new drug or sell more drugs, which are the things that pharma companies care about right? So you have to meet them where they are. Okay, what do people care about in PD and MS right now? It’s developing disease modifying therapies, right? You don’t want to bring these things to market, so unless you’re bringing data to them, that’s relevant to that—and that includes the labels that are coming along with that, and the longitudinal of the data and things like that—they might not care.
If all you’re doing is predicting a tremor based off of a subthalamic nucleus STN, they’ll be like that’s cool, we don’t have a lot of use for that. Now if you can link pathological oscillations to fast progressing versus slow progressing disease phenotypes and build a way for us to detect that as we’re screening folks for a trial, hey, that might be useful. And so I think there’s some dialogue in that. Sometimes the things that seem potentially useful are interesting at first glance or maybe novel, there’s still a gap. Maybe sometimes it’s just the heterogenetic patient population, sometimes there’s a scientific gap that needs to be filled before you can really make that marriage. And then that’s something that we kind of sit in and sort of facilitate sometimes.
But yeah, I think that sounds like that’s kind of what was going on in that circumstance. I mean, there’s pharma companies that are super hungry for data, right? I mean, look, if you look at on the oncology side again, just take Roche, for example, they paid about $3 billion for Flatiron data. They paid, I think when it was all said and done, like $6 billion for Foundation, again, mostly data. These are billions of dollars people were spending on…and I think it was Flatiron, but it worked out to somewhere between five and $15,000 per patient. They will buy, but it has to be data that’s useful to them. It has to be stuff that moves the needle for developing drugs or differentiating their therapies. Gaurdent has been doing this kind of cool stuff, it’s like companion diagnostic world. So, if you can be able to do companion diagnostics that again, differentiates your thing in the market. But yeah, just saying like, “we can do this thing with our data.” Maybe? Maybe it’s useful. Maybe it’s not.

Matt Angle:
Brian, you’re finding actionable on what timeframe is it something that could feed into early drug development or things will inform their decision making in the next two years? What do you finds striving your interactions?

Brian Pepin:
Yeah. So there’s kind of a two time scale. So one is what can you do in terms of interacting with clinicians today? Providing them information that helps them understand when a certain therapy drug, neuromodulation waveform, is going to be the right thing for that patient. So that’s very actionable. And you can think about. It can be a cost of sales thing, driving commercial activity.
And then the other place that I think we have the best interactions is fairly early on. So maybe when somebody’s coming from research into development or maybe phase one to phase two is a good time, where there’s still some flexibility in their data and biomarker strategy. Neural data should be part of this strategy, whether it’s informed by neural data, or…a lot more folks are actively doing EEG, neurodegenerative clinical trials right now, or various types of imaging. And for those who’ve been following the whole Biogen, Alzheimer’s drug, one of the interesting things about that is it’s going to double or triple the amount of PET imaging that’s done in America, assuming that it gets widely dispersed because there’s so much PET imaging required for the ongoing sort of delivery of the therapy and validation of it. So—

Matt Angle:
Now I have to stop everyone for a moment because we’ve had two mentions of the new Biogen drug, but we haven’t told the audience what it is. Could anyone say briefly what the drug and why the approval was controversial?

Brian Pepin:
Do you want the politically correct version?

Matt Angle:
No, absolutely not. I have no time for that. My understanding is two people resigned from the FDA as a result of this and that most of the editorial output around this drug is that it probably isn’t very useful. So I’ll break the ice by saying that I’m a skeptic. And then I’ll take away some of the political price that you have to pay for now giving us the blunt version.

Brian Pepin:
Right. So, this drug is called Aduhelm. It’s meant to target protein buildup in the brain for Alzheimer’s patients. The mechanism of actions of this drug is sort of based on a hypothesis, which has largely been rejected by the scientific community at this point, that removing these protein deposits is going to make Alzheimer’s patients better or reduce their disease progression. I think that was a hypothesis that was popular maybe 10 years ago? And through a series of learnings, kind of across the neuroscience spectrum people don’t, at the scientific level, really believe in it anymore. And that the trial results basically bore this out, like very poor, very low efficacy relative to placebo. But on the other side of that, there’s no drugs for Alzheimer’s available right now. And so, what you have on the other side of that is Alzheimer’s patient advocacy groups, who Biogen is close with, who were really pressuring their legislators and regulators to approve something very loudly.
And I say if you have a lot of very concerned patients and caretakers of patients and children of patients who are loudly advocating, “approve something,” it’s hard to say no. And so it got approved with a very low realm of scientific evidence. Which on the one hand you might say, okay, that seems fine, right? It’s a drug, they can study it. We’ll get a lot of imaging. I think that’s all very interesting. I think what’s caused an additional level of uproar is that Biogen has elected to price this drug at $55,000 a year for something that it costs them very little money a year to manufacture.

Matt Angle:
Is there any indication that payers will pay for it?

Brian Pepin:
Well, that’s the big discussion that’s going on right now. Nobody’s made a payment decision yet. And obviously the big payer decision will be Medicare because most people who would use this drug are older and in the US on Medicare. But I mean, you’re talking about potentially billions of dollars per year, that would be spent on a drug that may not do anything. Or it may just have a very, very, very tiny effect. That’s a big conversation. Now, everybody else who’s had a failed or borderline failed drug in a similar class, Lilly, now they want their drug approved.
And so you have a bunch of this activity of people trying to approve drugs that are based on a rejected scientific hypothesis. It’s strange, because on the one hand, it’s great to see something maybe approved from the patient perspective. Long term, if we’re approving things that don’t have any effect that I feel like somehow damages the credibility of the medical establishment, the scientific community certainly doesn’t do anything. I mean, it’s, it’s all going to pass through, obviously the taxpayers or healthcare premiums eventually in terms of money. So it’s expensive. Is that really what we want to be doing as a society? I’m not as sure. And I think that’s where you’re seeing a lot of the press come in as well.

Carolina Aguilar:
Perhaps input if you wanted on your previous question or you want to continue with this one about the drug?

Matt Angle:
No. I wanted to give you a chance to talk about strategics and especially, you spend some time at Medtronic and I think you’re probably thinking about this ecosystem and how to get people more excited and I’d love to hear your opinion.

Carolina Aguilar:
Yeah. So, so one thing on the pharma evolution, I also don’t have any doubt in my mind that’s going to converge. I’m seeing it already as we speak. And there are therefore drivers for that. I think one is that things are getting more minimally invasive, right? And then it’s easier to get to a bigger population cohort. And this is what pharma is interested in. Second, when we are now talking about targeting peripheral nerves, also the indications that you target with those targets are much bigger. We are talking inflammatory obesity, these bigger parts and the cost to develop a neuroelectronic therapy is a 10th of what it cost to develop a drug, right? So you start putting on that plus that the data, for instance of neuroelectronic therapies is real time versus drugs. You have to wait or you have to see, and of course they’re going to be interested right?
And I think they’re going to learn a lot on the biology to keep developing drugs for those targets. So I think that is a way for sure. And I think it’s exciting for the ones that are in neuroelectronic therapies, right? Because on neurotech in general, at the end we have more options to partner to integrate, to compete, right? The ecosystem is definitely getting much bigger. When it comes to strategics, that’s what I wanted to say–that I think the scope is enlarging and this is very exciting. And then you see also the data-driven partners, also getting interested in healthcare, which makes it even bigger, right? So, it’s going in all directions, pharma versus medtech, data versus healthcare.
And at the end you start making this a lot more interesting. And of course we have to continue to make it safe, variable, and effective of course, for the patients, but it is changing so much that I’m so excited that we’re in this decade, because in other decades, we will have a much narrower scope. But I don’t know if that solved the question that you wanted to know.

Matt Angle:
I agree with you that bioelectronic medicine, particularly Vagal Nerve Stimulation and things happening at the periphery, you’re seeing a convergence of what would traditionally be pharma markets, that now what are device markets. Do you think that will find its way to the brain?

Carolina Aguilar:
Yeah, of course, right? I mean, at the end it’s a system. It’s the nervous system. So we have to connect, we have to connect all eventually right? Central with peripheral data and aggregation of data will be away. And who’s going to drive that aggregation, all these industries coming together and converging or collaborating right? I mean, it happened already, Galvani was an example and it’s going to keep on happening.

Brian Pepin:
The way that I think about that—I think some of our partners think about it as well—is that if you’re a pharma company, you’re thinking about disease mechanisms and targets. Historically, you’re thinking about things on a genetic level, molecular level, protein level, cellular level. And yet, the endpoint for Parkinson’s, like a patient motor diary or like some score of cognitive decline.
And so there’s just this massive gap between what you can test in a organoid or a Petri dish or animal model and what your readout is in human behavior. And I think neurotech and devices are filling that gap in between. So they’re saying that obviously anything that you change in the brain at a cellular medical level is going to be mediated by behavior is going to be mediated by the neural circuit. And so having some understanding of what that looks like pretty essential for having any real links between clinical phenotypes and underlying disease mechanism. Otherwise, it’s just, it’s just a big black box, right? And that hasn’t served us very well.

Carolina Aguilar:
Yeah. But you start seeing the connections, right? I mean, now there’s papers about connecting within the central nervous system, the pathway from the deep and from the cortical. The other day, there was a paper that connected the Vagal nerves with cortex. I mean, at the end, eventually now there is the connectome project, right? The connectome is more central nervous system. Why not to extend it to the full body? That’s going to happen as well.

Matt Angle:
Kunal, you’ve been posting a lot about optical BCI lately. I’m curious, is there a future where there could be a therapeutic and Inscopix product?

Kunal Ghosh:
I think the journey for an optical BCI is probably a long one. It does rely on some gene therapy, not just device translation into the clinic, but it’s not impossible. I think if someone asked me five years ago, is optical BCI ever in the future of Inscopix or for the field that I’ve said no, but today I’ll say, never say never. I do think the explosion of gene therapy companies and gene therapeutic strategies also for CNS has been incredible in the last few years. And we, ourselves are working with at least one gene therapy company. So I do think that it’s not impossible, but having some said that, our role in this space is to help inform better stim, electrical stim, help inform more conventional neuromodulation either from some of the bigger incumbents, like the Medtronic or the Abbotts of the world, or some of the generation technologies that are going after specific circuits with much higher bandwidth interfaces and with the promise of more precision.
So I think for Inscopix in the near term, it’s how can the platform, the cell type specific imaging in macaque, for example, really guide better electrical stem, higher precision neuromodulation, and enable really overall better BMI development without necessarily translating directly the technology into humans in the context of an optical BCI platform.
But having said that, I think that is not impossible. Maybe a few years ago, I’d have said, yeah, I don’t see that happening, but it’s really incredible to see how much gene therapy has advanced.

Matt Angle:
I love what a diversity of companies we represent here on the call. And, and as a way of closing, I’d really be interested in hearing from each of you, where do you see a gap? What would you like to see the next neurotech startup be? Where would you be a big cheerleader for a new company?

Carolina Aguilar:
To me, it’s integration of, I mean, synergistic collaboration to get the most of the understanding of the brain and the setup options, we can deliver it with our strengths, but really that interoperability somehow needs to happen. And again, for that, I’m extremely convinced that the people development models also need to change or that needs to change. I don’t know if you have read about these companies or seen those more distributed leadership companies with mindsets that are not about ”I’m going to be famous” but you know, I’m going to create an impact in the world.

Matt Angle:
Brian, what do you want to see?

Brian Pepin:
Yeah. I mean, Kunal mentioned gene therapy. I think there’s going to be some really interesting stuff where folks are thinking outside of the normal boundaries and doing things like gene therapy where you’re targeting with some specific vector to cells to give them like a sonogenetic character and then using some external ultrasound transducer to activate these cells. So you’re getting kind of the benefits across multiple, these spaces, your cell type specific, but also location specific and also temporally specific.
I hope the work that we’re doing collectively as a this kind of generation in neuroscience and neurotech is kind of laying the data and neuroscience framework for some of those things to emerge, find some of these new targets that, that folks with that kind of expertise that can then go and build a therapy for. Whenever I go to like SFN or look at, I guess, last year looking at the posters, I kind of spend some time over in that section. Cause I think it’s just, it’s so interesting and so rapidly developing as well. The velocity is really exciting.

Kunal Ghosh:
I think, from my perspective, if we have better approaches to get data from human brains, of course, that is somewhat indicative of the—I’ll call it the circuit phenotype, right? So not in a super course, neither does it have to be super fine, but somewhere in between and the circuit seems to be the right convergence area of convergence. And the analogy I’m drawing here is, I mean, Brian mentioned this earlier, the ecosystem of precision medicine companies in oncology. I mean, of course there was 20, 30 years of microarray and Illumina fueled genomics research that all these companies are kind of riding on. And I feel neuro is where genomics was 20, 30 years ago. And then in neuro the arc is probably longer because the brain is much harder and where talking about a much more complex set of processes that in many ways are intertwined.
So there isn’t necessarily a clean “genotype” that underlies specific mental illness or disease function. There might be specific genetic predispositions, but we all know that the environmental factors also have a huge role to play in how disease is manifested. So it’s hard to draw a direct analogy to genomics, but oftentimes I’ve taken a step back and I’ve asked myself is there a parallel arc here where the kinetics might be longer? It might not be 20 years and in fact, I’ve heard Francis call and say, it’s probably more like a hundred years. There could really be a trajectory here where what we are seeing is the foundational kind of elements of an ecosystem where we are collectively figuring out how to decode the brain. We are coming up with more precise means to gather either preclinical or human clinical data that can then give us hopefully the kinds of applications that have since been enabled in oncology and personalized medicine based on precision genomics data.
So can we have the analogs that we have seen in oncology, for example, here in the brain space where we have—let’s just, for lack of a better term—have specific circuit phenotypes for Parkinson’s for different mental illnesses? And can these then help fuel development of companion diagnostics, the development of therapeutics, and that becomes a substrate for innovation and for clinical impact. And can we also then essentially end up having the same kinds of benefits that we are today seeing for a lot of cancers where cancers are becoming increasingly more and more manageable if diagnosed early, thanks to a lot of these precision oncology and companion diagnostic innovations and all I think fueled by the genetics genomics revolution that we saw powered by companies like Affymetrix and then of course, Illumina, and then a plethora of genomics tools companies.
So who are the right companies in this space, the next generation of companies that can accelerate the kinetics? I think those are the companies that I would love to help going back to one of your earlier questions; if we were ever to become investors, help invest in because I think the space needs more shots and goals. It needs more companies that are in it for the long game. And I really do think that there are some parallels to how genomics and now precision medicine is evolving for the field of brain science and precision neuromedicine in the future.

Matt Angle:
Carola, Kunal, Brian, thank you very much for taking the time today.

Brian Pepin:
Thank you Matt, for putting together this pub session. I’m drunk now. So, I have to go home. Two hours in the pub!