Intro - 00:00:08: Startups Podcast by Excedr. Join us as we speak with first time founders, serial entrepreneurs and experienced investors about the challenges and triumphs of running a biotech startup from pre seed to IPO with your host John Chee. Welcome back to the Biotech Startups Podcast. Today we bring you part three of our conversation with Scott Chatham, co founder and CEO of Faro Health. Last time Scott discussed his transition from academia to industry, his commercial leadership at impedimed, and the organizational frameworks required to scale technical teams. In this part we examine the massive hidden engine of global clinical development. Scott and John break down why clinical trials don't follow linear economics, the structural divide between big pharma development machines like Merck and cash conscious biotechs, and the staggering complexity of protocol authoring. Scott explains why human teams spend months manually evaluating protocol trade offs and why automated scenario modeling and AI evaluation represent the key to compressing clinical trial timelines by 50%. Let's dive into part three of our conversation with Scott Chatham.
Scott Chetham - 00:01:40: Yes and no. The reason I'm saying yes or no it's because it's nonlinear through the program and it depends on the nature of the company. Big companies are just development machines and they have fixed sitting staff and the more you put through it, the more efficient the system gets. So Merck is an amazing machine. Like just amazing. Like it pushes things through clinical development in a very efficient way. I mean there's a New England Journal of Medicine. What is the strongest predictor of a success of getting a drug to market? First one is IU Oncology.
Jon Chee - 00:02:11: Okay.
Scott Chetham - 00:02:12: And the second one is IU Merck.
Jon Chee - 00:02:14: No way.
Scott Chetham - 00:02:15: Independently researched. And it's basically because they have this big machine they've invested in the quality of too and and working with key hospitals and physicians to be able to have people pre contracted. So systems that are effective like that if you push more through them, the cost goes down quick. So if the bigger companies who really do the development so think of pharmaceuticals as like the research part gets done in biotech and small companies and the ideation and then you aim this, you sell to one of the big companies because they are development machines they will get it the last phase through, they'll get the registration, they'll get the ability to sell.
So in those machines yes there's a huge one to one you get down by 50%. If you pull it down you'll see a massive cost reduction because you can put another thing through smaller companies it's not that nice because they have nonlinear loading like you can shrink, they burn a lot of cash. Their most expensive part is when the whole team is hired to run the clinical trial or a CRO in the most likely. That's the most expensive part. So amendments cost a lot of money. Waiting around for data like interim data analysis costs money, so it's a less linear thing. But there'll be big impacts there because a lot of that will be automated, not necessarily by us, but by other companies doing that chunk of the stack.
There's a lot of white space, I would say, in clinical development for people to play in and everyone should do quite well. So I think, yeah, you'll see it. It's just going to be a little bit different. But for the most part, if this all works out, we should see a pretty hefty reduction in the cost of development. I personally like to believe that most companies will pass that on. Other more cynical people say they won't. But honestly, I've been in the rooms with the people and people are only in this industry because they want to help people. You'd be surprised. It's not an easy job. It's extremely difficult.
Jon Chee - 00:04:02: I was going to say there's easier ways to make a lot of money.
Scott Chetham - 00:04:04: Yeah. And you don't choose this career because that's what you want to do. And so I think it will. It might not be straight away and it might be incremental, but I think we'll eventually see like a reduction in cost. Because. Because this is why some of the genetic messes are like a million dollars. It's because they're personalized. I think the exciting thing for the melanoma vaccine, the brilliance behind Moderna is not, I think, what people think it is. So MRNA technology, MRNA has been around a long time. I'll just state it because I always do. Does not alter DNA. Just in case anyone starts thinking that
Jon Chee - 00:04:35: does not just the level set here.
Scott Chetham - 00:04:37: Viruses alter DNA. MRNA does not. But MRNA is very cheap and can be chemically made. That's the beauty of it. And then it programs cells to make nice like interesting shaped proteins that cause immune reactions. What Moderna has is this brilliant piece of technology that is like an inkjet printer that lets you produce personalized mRNA, which is why they can personalize a vaccine for each unique individual's form of melanoma. That industrial process to do that is ingenious. That's the real power and scale of that company. It's not them RNA like yeah, you had the delirium mechanism is like lipid and wrapping it. But what they really spent years doing is being able to mass produce unique vaccines at an individual level at a cost effective price.
And I think finally the market's rewarding them. Unfortunately they just because of politics has been a terrible ride. Unfortunately for those people, I feel really bad for them, so painful. They're good people but there's a lot of brilliance in what they've done and I think now that that data's out I'm really excited to see where that goes because again without that industrial part to mass produce unique vaccines, a lot of that next wave of medicines wouldn't be possible. And I think for long term investors in Moderna and I have no inside knowledge here, I just think if I could invest in pharma, unfortunately I can't because we have too many costs, it would be conflicted. I'm conflicted enough so I can't. But if I could buy shares in the donor I would have bought them and held them a really long time because I think they'll make it just in an unfortunate time. But the technology itself is really ingenious
Jon Chee - 00:06:16: and for companies that are like you mentioned, you have the kind of pre clinical work staying in startup land and then eventually you hand it off to get the clinical work done, commercialize, get this thing out the door. Do you think with all the technology improvements, cost efficiencies that we're starting to see maybe in the next five years that we will have companies that become the next Merc where we have more than just one engine like one clinical engine that's just humming like is there going to be the next, like a next iteration of that?
Scott Chetham - 00:06:47: I think we're going to get the commodity machine of being able to get good development processes and everything set up will probably become commoditized and I think we'll reach a fairly level playing field. I mean some are just going to be always a little bit better because they have the right relationships and other bits. But I do think the strategy of how you get a drug to market, the choosing what we call an endpoint, like what is the thing you're measuring and who you're measuring in and when. That's now the key differentiator for all of these companies. And almost your entire success lands on those decisions. And that's the art what I see and this is kind of what we're launching product in. This is you can automate the research of that you can automate the synthesis of collecting vast amounts of historical information. Everything that's ever changed before for you can automate like all of these things, it still requires human expertise. Is this the right endpoint? I'm going to measure and at what time, point and then how many. And you can use LLMs to like, we do it to like estimate it. But in the end, there's art there and I don't see that going away. That's some of the most important knowledge that sits in a lot of these companies is like, I'll go back in history a little bit.
I'll use Amgen as an example because I've probably spoken about Merck too much. I love Merc. I love that company. Great people, really good at what they do. Amgen, like, for osteoporosis. So unfortunately, those drugs are off the market for other reasons, but they became the leader. And the reason they won is there's no objective evidence that the rival molecule that came up a little bit later was any better in reality. But what happened is Amgen chose to measure it first in a smaller number of people, fracture risk and go for a smaller chunk of the market, improve that smaller entry point and get an approval by doing that. Then they kept their clinical trial running and enrolled people of a different age group at a different risk profile, and they would expand on it. Where the rival went something more like, well, I'm going to try and capture the whole market at once. And they failed. And so Amgen, through better just decision process and better planning, in some ways pipped the other company. Because also now when you get an approval, the next drug has to be superior to yours or equivalent. And that's a different trial design. So you've kind of screwed them.
And so there's a lot of art here. So what I'm kind of getting at is that I think some of the actual process and the machinery is going to get more commoditized, but the art in what am I measuring, who and when is not going away. And that's like human judgment. That's market analysis, you heard. It's also competitive intelligence. It's also like you can automate all the work up to that point, but in the end, some people are going to have to make that business decision and that's not going away. Could you ask an LLM for a decision on that? Sure. I would say it's probably worse than a human in this particular case, because we can kind of make bridges to net new things better. Because keep in mind, the other risk is like with LLMs and other bits and pieces, you're trained on all the data in the past. We're inherently probably doing net new biology. So there's a certain amount of projection and estimation and other things outside of any governing data model. And so you gotta make assumptions and some of that is human judgment. And who knows, maybe we'll come up with aggregate general intelligence. That's better. I don't know. But I do know that automation of all the collection of that work, because it's what I used to do, takes months and now you can do it no time. That's a huge saver. That's the win right now.
Jon Chee - 00:10:11: The way you described it is really interesting because it's also terrifying the way you described it, because here's the thing, right? You start your pre clinical journey, tons of money going in, you're ramping up for your clinical trial and you have one endpoint to rule them all. And this thing like you see what happens like in the public market, like one bad readout could just be the end of you. It's like everything on the line, it's just like.
Scott Chetham - 00:10:34: And you are 90% chance of failing.
Jon Chee - 00:10:36: Yeah.
Scott Chetham - 00:10:37: Like if you want to know the probability, it's about 10%. It's getting worse by about a percent every year or two. We're going down, not better, unfortunately.
Jon Chee - 00:10:44: Is it because the problems are getting more complex?
Scott Chetham - 00:10:46: Yes, it depends on who you talk to. You get a different answer. The direct correlation is clinical trial complexity as directly proportional to like the amount you're measuring in people, your successful outcome. Some people tell you, then we'll fix the complexity. And there's an aspect of that that's very true. The flip side is it's more complex because the biology is more complex. So some of this balancing act between those two concepts, it is a very hard job. And you are betting a lot on the line. We do trials because we don't really understand biology. Not really. That's why we have to do them. And so you have a lot on the line. And who knows, maybe the amount of money that's plowed into like AI based design, molecule design, protein design, better targeting, will that mean we have less values for efficacy and less likely for safety and blah blah, blah. I hope so. But I think it's a reach to say still at this point in time that we can't see the impact of that. And I don't think we're going to for several more years. But the problem has never been having stuff on the shelf to study. We've always had more stuff on the shelf than we can handle. Our problem is the clinical, the funnel to shove things through this 12 year development cycle and the amount of capital it takes is what's the restriction? And it's one of the comments I keep telling VCs all the time. You've planned all this money into discovering more molecules.
Jon Chee - 00:12:07: That's not the problem.
Scott Chetham - 00:12:09: Jumped in. But unless we fix development, it won't matter because they'll stop paying all these fancy licensing fees because they can't study them. And I think that everyone's waking up to that fact. But the system is inherently constrained. I mean, yeah, we can get it down in time. That's my whole focus. But you still need a patient. You still need, like, the hospital. And it's a very human problem. This is where, I mean, if you're an AI company, I will tell you the human reality of it, of consenting a patient into a trial that has risk, because I've done it, is not something you can automate. It is a relationship. You have to have trust with the patient. I consented mail into a clinical trial and then I got pulled up in the hallway of the hospital and yelled at by his wife because she wasn't around. And he didn't say, I need to talk to my wife. And she said no. And so that turned into a whole thing.
So it's just complicated. And I think this is where faro I'm very much focused on the things in our control that we can do. I think all the noble effort for companies trying to focus on recruitment and efficiencies through that system, but it's an inherently uncontrolled system and it's a very human problem. Like, yeah, you can push a remote consent form to someone and they e consent on an iPad. Yay. Someone still had to sit there and talk to them about it. And someone still had to convince them that this was good for them. And not just them. Often their family and sometimes extended family. And then when the extended family calls, having a conversation with them. And so it's particularly for complex things like oncology. And there's only so many people. Like, I know we only 5% of people qualify for clinical research today, which is terrible. But it's hard to move that dialogue. It really is. And I think I'm very much focused on what I can do because I know we can fix my part of the problem and it will have a big impact. But someone's got to have to. Like, there's another chunk here that's just very human. Unfortunately, that makes a total sense.
Jon Chee - 00:14:02: And I guess a question for you. So when you were describing the Amgen strategy of like, starting off a bit smaller, you know, rather than just like, let's go boil the ocean and just like, try to win this in one go. When you're doing clinical trials and like, endpoint design and trials design, can you pivot? Like, we talked about the whiplash of like, just like pivoting. Can you, like, pivot in a clinical trial where it's like, not absolutely just like, okay, the readout wasn't great. We're cooked. We're now shutting it all down. Is there a system or a way that you can pivot and still get it through the clinic?
Scott Chetham - 00:14:34: Yes and no. It requires very clever planning. So there are things called adaptive trials. So sometimes you don't know the responder effect. Like, because the response effect for drug is much less than what people think, by the time you average it out between everyone, it's a lot less. And so you might not know that going in. So you might need a hundred more patients to hit it. That's a planned adaptive strategy where I can do that. Cause I planned it up front that I'm going to do a test and if I was close to, I'm not reaching statistical significance, but if I enrolled a hundred more people, I could get there. If I plan that a prior, I can do it. You can change an endpoint in a study. You can maybe change a primary and a set, you can maybe flip them around, but you'd have to collect more data. And you can do it while it's blinded. Like, I still don't know the results once it's unblinded. That's phishing. What some trials do, and it's worth it, depends on the commercial strategy and all sorts of things. You can run two trials in parallel with different endpoints, and you start one many months before the other. And the aim is, is you get a first read on the result and then you have a second trial that's running like six months later, whatever timeline, and you don't know which main primary endpoint you're going to go after or something. And then you see the readout of the first one, you go, oh, oops, I'm going to flip them in that one. It's still a guess, but you could flip them because it's still blinded, still blind.
Jon Chee - 00:15:55: Got it.
Scott Chetham - 00:15:56: So I can do that scientifically without breaching anything, because I now know from an earlier trial that that was the wrong approach and that's actually led to set lupus drug approval. Several things like, that's a valid strategy, but it all has to be planned upfront. So a lot of what you go into, and this is where AI can help stuff coming from us in the future is it takes humans months and months to pull all this data together. That's the real work. It's painful. You have to sort through papers, you have to sort through publications, pirate stuff. You can automate that harvesting and present it to people, all traced and verified. That lets people then consider the full landscape in a way that is very hard to do. Think of it as like legal tech, legal technical precedent and all these other things. Same thing, same reuse of the same technology. Like similar stuff is kind of what they're doing now for some aspects to help people know what precedent was and get it information they can't find on their own and also go down because of embeddings and things. You've got a knowledge graph that can link concepts humans may not pick up. So it's a better, in some ways it's a more rigorous researcher than we are.
But then someone has to assemble it all and make sense of it because it's net new biology and there's guesswork and there's some estimates. And this is where you'll need like one of the four, like thought leaders. And this is where the human blends with some of this automation. And it's pulling that together in a beautiful experience that's collaborative and like that. That's the challenge. And some of the stuff like is said where we're going now is that is because it's hard. You know, it's what I used to do for a job and it's partly like I always wanted to do this. You have to start further down because one, it's where the value proposition was and two, the technology wasn't there to do this until now. You couldn't do some of the stuff we're working on even six months ago, but now it's there and so it's getting this stuff done and out previewing. Our customers are excited because we kind of show them what's coming and like, well, I want it now. It's like, oh no. Everywhere. Because we like to partner with them and get feedback early, which is a good and bad thing. It means I get excited and I want this thing now. It's like, yeah, it's not ready now. Like we're showing you controlled examples like make this stable and things.
Jon Chee - 00:18:02: That's super funny because that exact experience, I'm not a coder by any means, but because of everything that kind of is now available in the past few months. I can like spin up prototypes like pretty quick. Like I use technology a lot so I kind of know what feels good. And so my wife is in software, so like I see her running demos. So I played pretend for a little bit, a little pretend demo. And I was like, all right, let me just show you something I've been working on. These are like Fortune 10s and they're like, can we get in now? And I was like, nope, you can't. Like, this is a prototype that yours truly created. Not ready for you. Not ready for you.
Scott Chetham - 00:18:43: Well, I think that's what I love about this time and it's exciting is like I can now get my, like I use Claude code. Not a code, but I use Claude code like you wouldn't believe. And it lets me get my thoughts out of my head, get customer feedback, play out scenarios that didn't work, then get it into a concept, right? And then hand it to engineering and product to like build the stuff around it. And that compress, like would choose to take six to nine months you can do in like, if I had more time, I'd do it in a week. My problem is like, oh, I have a full time job and so like between things. But I think that's the exciting thing.
Now this is a bit of a pivot and I don't know the answer to this, but I've read this several times that I keep reading comments from VCs, like, the current generation of founders is the best anyone's ever seen and it's been publicly said many times now by a number of big VC firms. And I have a bit of a pushback. I actually think the fundamental difference is the fact that now you can prototype and get customer feedback in days. That cycle used to be like months to a year. And the reason we're seeing things move so fast now is because you can get that flywheel moving really, really quickly and you just couldn't do anything at this speed before. And so I questioned that narrative quite a bit. I don't think that's true at all. I think the tooling has fundamentally changed and I think a lot of like founders, obviously without legacy infrastructure and other things, you would just jump straight on this. Whereas like an older company like Faro, we had to pivot this company into like full spec based development and move QA processes took a year. We're not very big, but we have existing customers, we have existing process. That's not. But now we're all in that way and that's how we do prototyping. And so I just don't think that narrative's correct. It's just interesting to hear it.
Jon Chee - 00:20:28: Yeah. At least for me. And being in San Francisco, I'm like, just seeing all the kind of, like, startups, like, kind of like cropping up and bubbling up. It also seems like the age of the founder is, like, way younger now. Like, way younger. I was like, whoa, you're already through, I see. And you're like, still in college. I'm like, I was like, drinking beers with my buddies. You guys are like, way ahead. But, you know, jokes aside, like, just like toiling away at this company for so long. And I think you might have mentioned this, like, earlier, like, I'm just like, coming in with, like, an opinion on, like, these experiences I've had for building this company, working with customers, working with all the stakeholders for such a long time that I was like, trying to figure out why were people just, like, so enthusiastic about this, like, shitty replit demo that I made. And I was like, oh, it's because, like, I've just, like, lived and done this by hand and just, like, got feedback for so many years that, like, out of the gate. It's kind of just like, I can, like, exactly. We said you're like, tweaking with cloud code for us. We're just like, I was like, building in replit and just like, polishing this thing based on, like, my experience through building this company. And when I put it out in front, this is like, ah, yes, it resonates because this is like a real lived experience versus, like, you don't have this experience or like, firsthand knowledge. You might be kind of what you said, building a hammer and looking for, looking for the problem.
Scott Chetham - 00:21:46: You're looking for the right nail. Yeah, it's interesting. We've created a lot of competition. I think it's very hard to understand this field if you haven't done it for a lot. If you haven't actually done the work. It's almost impossible because it's so specialized and it's so siloed, even inside these companies. And I think building a platform here is just inherently difficult if you don't know the domain. And I think that I see cool things by some of the founders. It's not to dismiss them. They are cool. It's like, oh, someone was like, I'm going to borrow that. That's a feature. We'll have that in three months from now. Fair's fair. But yeah, this is just a very esoteric field. And I think a lot of the deep domains will stay that way. I think it's even going to be extremely hard for the big AI labs to come into clinical development. Not because they're not full of smart people, it's not because they couldn't if they really wanted to. It's more that it's highly regulated. And because of that, there are controls on what we do and how we do it, which would require them pivoting other parts of their company. And so they're not going to do it. You won't see someone in my domain then also sell like cart, like do something. You can't be in both. This is a very specific vertical and I think there's going to be a lot of space within. There's a lot of other examples. I think in other industries that they're very specific verticals that will probably stay quite protected just because of the nature of the business knowledge. And the data required to build us will never be in the public domain. And so we've got it through accumulated working with the big companies. It's the only way to get it. And it's interesting time, but who knows?
I did not predict when I said I started Faro, it was because I was like, and I will automate this. I just didn't predict it would be LLM and when I thought it would be sooner than something else. Working with the Brain team because we spent a lot of time on image recognition and a lot of other things, it was pretty cool. But then when LLM suddenly came around, actually the first time I saw the first one for Open, I was like, what a cute toy. It took me the second release, I was a bit slower. Soon, I think it was their second version. When it was like, the first one was like, this is stupid. When I saw the second one and how fast it came out is when I started to go, oh. And then by the third one, like their third release, I was like, oh, this is 100%. This is the future. This is where everyone's going now. So I was a little bit slower to it. It was the first one. I really idiotically, I did not see the future. I was like, oh, this is cute. Oh, I completely missed it. And it's so also because one of our VCs, like, yeah, you're making a huge mistake here. Dylan Reed of Z Adventures, he saw this coming. I'll give him complete credit. He pound me on it for a while and he was right. But it took me a bit longer than others to see it. And then, yeah, then I got there very quickly.
Jon Chee - 00:24:23: Yeah, yeah, that's exactly. I was a bit of a late bloomer as well, but now I'm like, I like definitely can appreciate it. So like, you talk about all these strategies for design of clinical trials, enrollment and all of this. Does it add a layer of complexity? I'm guessing these like multiple trials are happening globally too. So like I'm imagining trial design in United States very different than the eu, very different than in Asia. And when you're doing a clinical trial, planning this out up front a priori, are you also like doing this globally?
Scott Chetham - 00:24:58: Like, are you like, everything's done globally? Predominantly when you get to late phase, because you're selling to the world, but you have to be very careful in the countries you select. Like for example, right now you'd be cautious to do any work in Germany because of the new regulations on most favorite country pricing. And Germany charges less than everywhere else. Like the cost of drugs they buy at a fraction of the price. They're brutal in negotiation. So if I was a pharma company and I sold to Germany in a trial, it would set the floor on my price for the rest of the world. So I just won't go there. So the system is going to change in a big way because of a lot of the changes in government and regulations here. That is like now cascading through the world. Because why would you sell to that country? If I now capped my pricing worldwide, it destroys me. I can't get my 20 billion back for all the failures and so I'm not going to go there. So I'm just never going to sell in that country. So it's commercial and so you have to plan this out.
But today it's all done manually. The future in the perfect world would be planning would be so fast. I could create scenarios for every country and run an optimization algorithm and I'd pick the right one. No one can do that. But that's the future. That's the future. That's where we'd love to head, where we're heading. I'll be blunt, but it is a capability. It's kind of in some ways the holy grail of where we need to head as an industry to be much more clever on the sequencing because this takes 12, like 12 if we get down to 6, which we should. I still have to run the back to back. I run early phase research where I'm is this dose safe? So before I do anything, I do give healthy. Generally, let's say I'm not in oncology. I can't give those drugs to everybody. There's some drugs that are just toxic. So for non toxic drugs I would give healthy people small amounts and escalate the dose. That's a phase one. I don't care if it's effective. I just want to make sure I do no harm. And so I do a lot of safety. Pre, you know, they're done in specialized centers. I do a lot of safety. I look at organ health like biomarkers and all sorts of things. I escalate it up. Very specific lock designs. Like I can only enroll a person at a time. If I'm in three countries, I have to share the data between everyone. No one can enroll the next patient till the results come from the first one to make sure it's safe. Like there's a lot of interlocks.
Then you go to phase two which is when I try to get an idea of efficacy and I'm in my disease population that I'm trying to treat. So the first one is am I just safe? And the next one is like am I measuring the right thing? Can I give this population I'm going to study, can I give enough of this dose to see an effect? Because I don't know how much to give people. I only want to give just enough to see the benefit and nothing more. That's a phase two and so that's a specific design. Then I'll do that's a phase 2a. Then there's a phase 2b. I choose a dose. I try to get an early efficacy signal. Is this thing working? If I see that I go to the big phase three, those are the huge ones you hear about in the news. If they foul ass like very expensive, like insane amounts of money. Being able to optimize through that, which countries each step happens in. Like what I run in parallel. Do I go for a smaller indication first, like notice for GLP1s. My first one indication was weight loss. So I was type, treating type 2 diabetes, then weight loss, now a reduction in cardiovascular disease. There's a label expansion. So as you do that's all strategy. And so it's, that's putting all that together just takes time.
Jon Chee - 00:28:05: Every like phase you just described global
Scott Chetham - 00:28:08: At every phase it depends like phase ones can be. I may only have 10 to 12 people in a phase one, but I may need to spread it out. You also have other issues. But I mean planning that strategy is just a lot of work. That's where I think ML and AI will be very helpful because it can do what each scenario looks like and tell you what it will cost and like how long it could take to enroll. And then it comes down to like you can run optimization algorithms into all sorts of things that would let people see the landscape in different ways. How do I compare against my competition? So I think we talk about AI design, but in some ways in these scenarios it's like being able to do run all the permutations because we just can't. Humans can't do it. Just to do one alone is like three to six months of work for a human.
Jon Chee - 00:28:51: Oh God goodness.
Scott Chetham - 00:28:53: To do it properly. I think good companies is three months of multiple people. And that's not to say six months is bad company. It's more that some things are just more complicated.
Outro - 00:29:04: That's it for part three of our series with Scott Chatham. Next time Scott recounts leaving Google and verily to launch Faro Health, navigating the shocks of COVID and the generative AI Wave, and his candid perspective on late career entrepreneurship, capital, efficiency and customer trust. Join us on Thursday for the final part of our conversation with Scott Chatham. The Biotech Startups Podcast is produced by Excedr. Don't want to miss an episode? Search for the Biotech Startups Podcast wherever you get your podcasts and click subscribe. Excedr provides research labs with equipment leases on founder friendly terms to support paths to exceptional outcomes. To learn more, Visit our website www.excedr.com. on behalf of the team here at Excedr, thanks for listening. The Biotech Startups Podcast provides general insights into the life science sector through the experiences of its guests. The use of information on this podcast or materials linked from the podcast is at the user's own risk. The views expressed by the participants are their own and are not the views of Excedr or sponsors. No reference to any product, service or company in the podcast is an endorsement by EXCEDR or its guests.
