Noam Solomon - Immunai - Part 3

Intro - 00:00:01: Welcome to the Biotech 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, Jon Chee. In our last episode, we spoke with Noam Solomon about his decision to pursue postdoctoral studies at Harvard and MIT and the impact of being surrounded by brilliant minds. We also discussed his transition from academia to entrepreneurship. Pivotal moment that sparked the idea for Immunai, the complexities of AI and ML, and how Immunai is trying to revolutionize drug development and diagnostic testing by leveraging the power of data and machine learning. If you missed it, be sure to go back and give part two a listen. In part three, we talk with Noam about approaching precision medicine in an entirely different way and the challenges he faced in the initial stages of establishing a lab in New York. We'll also discuss the intricacies of working with pharmaceutical giants, the delicate dance required to navigate these relationships, and the importance of fostering a culture that embraces failure as a learning opportunity while avoiding catastrophic missteps. Lastly, Nome reflects on the evolution of Immuni's approach to innovation and collaboration in the pharmaceutical landscape.

Noam - 00:01:37: I didn't come from Bialyber. And it didn't come from management background. And when we started the company, it was a very naive mission to kind of find the right therapy for the right patient, to help patients that have very difficult diseases. There was no business model in mind. There was no understanding of what are the business unmet needs in the space. But maybe I'll take a step back and tell you something that I think was very interesting for me as I was kind of studying the subject, that in the early 2000s, the Human Genome Project was completed. And back in the day people thought that, okay, this is going to finish all of our diseases, right? We're going to now, we understand the DNA, people are going to now get kills, you know.

Jon - 00:02:27: Yeah.

Noam - 00:02:28: It did not happen. And you know, far from that. I think there are certain diseases and certain treatments that happened because of the mapping of the genome, but it was just the beginning. And then there was, and I view this as the, kind of mapping the alphabet of our genetics. But you know, the alphabet, you can't read books only with the alphabet. You need words, and then you need, these are the cells in our body. And then there was an effort to map the cells called the Human Cell Atlas Project. And this is still an ongoing effort. You know, it's aiming to map the cells in our body. There are many cells that I mentioned, trillions and trillions of cells. But this is like the words in a book. The words are not describing the narrative, the story. How do you, you know, read words in a book? And this is like... In biology, it's called systems biology questions, right? In our case, we are trying to map the system question, which is how the different cells interact with one another. What is the story of it? What is the narrative of the story? Why are different people have different narratives? And so it is clear from the way that I describe it. To map the human genome was a fairly simple exercise. At the time, it looked like a very complex one. Then the human cell atlas, it's still complex, it's not finished, but the human immune system, which is the book of the immune, of the immune, and there are, you know, billions of people around the world, they have different immune systems. This is already a very complex problem. Nobody. I think seriously try to actually map the human immune system. People, some companies and some researchers try to map certain mechanisms in the immune systems. For example, to map the T-cell receptors, you know, the immune repertoire. This is something that has been tried. And it's interesting that even today, many years after it was not completed because we don't really know how TCRs bind to antigens. They don't want to get too much in the way, but there are certain problems in biology that are still wide open despite, you know, having been studied for 20 years or so. When we started, most companies were Thinking about treating patients by looking at Let's talk about oncology in cancer. So cancer patients will treated by understanding the mutations in the tumor, mostly the DNA mutations in the tumors. And the thing is that different people have different mutations in their tumors. And also the immune system is the common denominator for everything that we encounter. In your life, in my life, we have covid and we have the flu, we are aging. And I felt that this is like a crime story that you are only studying the perpetrator, the criminal. What about the victim? So the human immune system is being now assaulted and nobody really tried to understand it because it was more complex than studying mutations in the DNA. So I think that the benchmark or how people or other companies were thinking about this problem was to go and understand what are the mutations that are possible in breast cancer and then trying to find diagnostics and treatments for these mutations or for patients with these mutations. We were looking in a different angle that I think, I'm not saying we were the only ones or the first ones to figure it out, but we were definitely, it was a blue ocean at the time.

Jon - 00:06:08: And this is incredible because I think you mentioned or might have called it out as like maybe as an outsider, it was like a naivete to it. But I honestly think that's a superpower because when you have an outsider's perspective, like if you're an insider for too long, you'll start to think like this is an infinite problem that can't be solved, right? But you came and you brought your math lens and it's like, no, it's finite. This is solvable. I'm not saying it's getting solved tomorrow. That's not what I'm saying, but it is solvable. We can map this. And I think that's amazing because when you bring this diversity of thought, different type of disciplines coming in, everyone will approach the problem from different angles. And that's how the best work happens in my opinion. So, you know, for us, Excedr is an equipment leasing company and I'm a former biochemist. I don't know many biochemists that dreamt of going into equipment leasing, but I came in knowing nothing. And I like to think that the bench scientist in me kind of approached the financial problem from a different angle that makes I think are special and solving a problem from a different angle is like, that is a secret sauce to that. So I love hearing that, to be honest. And so now you and your co-founders like, okay, we're now tackling this problem. Can you talk a little bit about your technology will make it special and unique, um, and how you guys are approaching it.

Noam - 00:07:39: Yeah, definitely. And maybe I'll say that, you know, Luis and I both came from, you know, the more computer science side, but very early on, like months after we decided to go after, we approached scientists, like top scientists in immunology and pathology, and we brought them on board. The first one was Ansu Manspathy, who was a very young researcher, principal investigator at Stanford. He loved the mission statement. He was, you know, a rising star, a pioneer of using new technologies in biology. To study immunotherapies. And then he helped us identify Danny Wells, who became another scientific founder, and then later on also Dan Littman, who is a very established immunologist. So it wasn't, I mean and I wanted to stress this because it wasn't we thought, okay, a couple of mathematicians and computer scientists are going to solve a problem in biology. I think we had I told you that, you know, maybe some politicians are humble. And so I think we were humble about, we were not going to solve the problem, but just applying, you know, AI in biology and it's going to solve things. No, we have to bring domain experts. We have to create a language that is going to be a new language where immunology and computer science are going to coexist and are going to fit, you know, each other. And so I think this was really the environment from the beginning. As you mentioned, I think not only my naivete, but my stated lack of experience and knowledge really drove me to ask a lot of questions and be even. I would say even proud about not knowing. So I think that when you have a team of experts and the new team member comes and his boss or her boss, they are experts, sometimes it develops a culture of like people being timid about asking the stupid question in the room. And I think when you see the CEO asking the stupid question in the room, you feel much better about it than it became. It became something that I'm trying to really nurture and reinforce. I think to me that the thing that makes scientists, especially in life sciences, but I think all over and also makes for good managers is just you have to be able to ask the right questions and inspire people or motivate people to try and answer the question. You don't need to answer the question for people. And I think not knowing if you are not mad about it and if you are humble about it, it is a very good setup for this. If you can bring people that are smart enough, autonomous enough, you're giving them the space to operate and they are brilliant, they're going to figure it out. You just need to create the environment for that.

Jon - 00:10:37: Totally. And that's exactly it. It's like when it, I think that's where real innovation, like you almost get to first principles and you're just like asking, why is it that way? Why is it that way? You start peeling back the layers of the onion to the point you're just like huh, and then the domain experts might be like, huh, like, you know, I didn't ask why here, you know, for probably right reasons because like, you know, all of my colleagues were like, there's, this is, this is not gonna, you know. Too hard. Too hard bucket. And you're like coming in, you're just like, I'm happy to ask the quote unquote dumb questions. That's like, I think a great way to lead and really, really fosters the. The openness of collaboration and exploration of any particularly hard problem. So as you and your team, you're bringing in domain experts, you and your co-founder, and you're basically setting up the Avengers. Can you talk about the technology you then go on to develop and what makes it stellar and stand out in your opinion?

Noam - 00:11:42: Yeah, so I'll tell you a couple of things that happened. So we started like January 2019. And I think the original idea was to go and work with hospitals and try to have their physicians, organizations to give better therapies. Really quickly, we understood, especially in the US, that there is no business model for that. Like hospitals don't have enough free cash. They are not incentivized in any way to improve the value of patient care. I hope it's going to change, but right now and back in the day, it wasn't. And so It really led me and us to think about like, okay, so what then? What do we do now? Who are we selling to? And then we realized that the same questions with some small adaptation are very sought after by pharma companies and biotech companies. So the same questions of knowing which patients are going to respond to a drug or being able to find the right dose or being able to find which indications to give the drug in. These questions, pharma companies are wasting, really wasting. Many billions of dollars every year, hundreds of billions of dollars a year. Just without any sort of data-driven solution. And even before we opened the lab, we knew that if we were able to build a machine to create from a biological sample of a patient, a human immune profile, what is their human immune profile, and then monitor this before and after they get the therapy and foreign companies would want to work with us. So we already knew that. And we also knew that principal investigators in a hospital, not the hospital, but the principal investigator in, you know, Harvard Medical School or Memorial Sloan Kettering. They don't have this tool. They don't have this, you know, I'm calling it the MRI for the human immune system. So they also have grant money they're going to pay us. And very quickly, we realized we have to build a lab. To take the biological samples that, you know, pharma companies and medical institutions were going to ship to our lab, and we will take it and we will use to the single cell sequencing. We didn't know how to do it. Again, we did not have the experience. I remember that some investors told me that I'm crazy being a mathematician to go and become a, what are you, a wet lab scientist? How are we going to do it? And I knew in 2019. Not going to take too many years. For companies that only offer computational analysis to be outdated because of tools like ChatGPT and others. I knew it was coming also to other disciplines. So I was thinking about how can we build something that is going to be proprietary to us? And if we can build this machine and we will be able to create our own data sets. Then this can create something that is very unique and differentiated and despite a lot of convincing by external parties. Me and the co-founders and the founding team realized that This is what we need to do. We opened the lab in New York City. And very quickly we signed the first project with the pharma company and then the second one. But I think that the idea was to do what's right, even if you don't know how to do it. And we hired, you know, two postdocs from like good universities to go and open the lab. We hired, you know, a lab operations person from a hospital and we just started. And it was hard, but it proved to be the best decision that I've made.

Jon - 00:15:35: And I love that insight where there's the two insights that I'm kind of drawn on this. The first one is that you kind of your role as CEO. Is to look around corners and like, that's, you, you gotta like look ahead for the whole company and you saw it coming. You're like, how are we going to differentiate here? Like, and that requires setting up real physical infrastructure, but The other insight too about that is that I think and startup, generally speaking, and maybe this is a kind of a trend that I kind of see for the past, you know, 20 plus years is that. The capital intensity is always makes people kind of like their pH and their stomach turn like for investors. Too much time, too much money. Why can't we just be a capital-y light? Vertical SaaS company, right? And, but you leaned into like, despite the kind of perhaps pushback you're getting, you lean and you knew that this wave was coming and you needed to find a different angle. And despite it being hard, much like, you know, mapping the immune system hard, but we're doing it because it's important and it's a problem that's worthy of solving. You rose to the challenge and you're now, and like you said, it's like one of the most important decisions you made, which is awesome. And there's something to be said too, is just like, you know, you're like, I'm not doing it on my own. Like I'm not the only person who's going to be standing up a lab. Let's bring, let's like, again, using the Avenger analogy, let's bring people in who know how to do this and have stood up labs. We're going to be fine. We're going to be fine. Like they know how to do this. It's still going to be hard. I'm not saying it's easy. And you mentioned something, you know, I don't know chronologically how quickly it was, but you started signing, you know, partnerships with Big Pharma companies. Can you talk a little bit about just your experience and you and your team standing up that lab in New York and securing and educating Big Pharma on your product? Because I think those are, for anyone out there who's one trying to start, there's many startups who are trying to get their first lab up and running, who maybe are graduating from an incubator. And they're like, okay, you leave your parents' house. It's like, good luck like, go sign your first lease and do your thing. And so, one, that problem. Two, how did you approach getting your first set of customers and educating someone like Big Pharma who have been doing drug discovery a certain way for a very long time?

Noam - 00:18:14: Right, so maybe I'll start by addressing a point that you have made. We could have very easily become a competition company that only analyzes data received by the clients and partners. I think this would have made us a much smaller company than we are today and really make our potential upside or like holy grail, much smaller than it is today. And I think there is something about being a CEO that you receive a lot of advice. And you have a lot of advisors. I really think it's important to have good advisors, but it's more important to know which advice to take.

Jon - 00:18:56: Yeah.

Noam - 00:18:56: And some people don't like to listen to advice. I think that when you don't have the background or the experience in a specific field, if you don't have people that have the experience and can provide advice. You're going to really be naive all the way through. So the question is how to combine the naivete with the experiences that is out there. And that's really where I believe that critical thinking and the mathematician in me really helped me weigh in different advice and then take the right decision. Sometimes maybe the wrong decision, but being able to make a decision after hearing a lot of. Different opinions. And I think in this case, we made a very important decision. And I'm so happy we made the decision. And you asked me the question about the technology. So the technologies that we are using is first, this MRI for the immune system is the ability to turn a biological sample, either a peripheral blood sample that is essentially the collection of white blood cells, and then profiling the mRNA molecules and the cell-based proteins and the TCR and BCR. So it's called single cell multi-omic. So you really measure on the single cellular level, a lot of different omics, the genomics, the proteomics, and you create this huge metrics of like many, many cells. And for each cell, you have tens of thousands of measurements. So that really, really... High dimensional entities. And you measure them before and after being treated for the disease that they have. And then you want to look for the differences between patients that respond to the drug and patients that don't respond to the drug. Or, you know, between patients that have a severe toxic event and patients that don't. And this is a very complex question, a very complex question. In biology, it's more complex than in other areas. But I wanted to say something that you know, sometimes people in biotech, they don't appreciate, I think, the background that people coming from tech have. And I wanted to mention something, which is what is the fundamental question that a precision medicine company or a diagnosis company have when they are trying to know whether a patient is going to respond to a therapy or which patient will respond to a therapy. You can think about this as building a matrix. Well, the rows of the metrics represent patients. So you could have maybe thousands of patients. And the columns represent the drugs. Oh, you know, you have different drugs and you really want to know which patients are going to respond to the drugs or to assign the drug to the right therapy for the right patient and 15 years ago in a completely different industry. Netflix post, the Netflix prize or the Netflix challenge that ask people to come up with algorithms that will allow them to know which movies you're going to like, right? So not which drug you're going to respond to, but which movie you're going to like. And think about this at the intuitive level. In 2007, 8, 9 people say, you know, it's all a matter of taste. How can you know? Like, your taste is not like mine. But at the end of the day, by knowing a few movies that you and I like together, maybe there are some dimensions that make us very similar. We are not the same, but we are similar in these critical dimensions. And this opened really the field of recommender systems. And when I started the company with Lewis and with my scientific co-founders, I thought about this problem as a recommender system problem in drug development. And so I think that on the background that people from tech bring or people from one discipline bring to another, sometimes it's very important. Other times, maybe it's not so much.

Jon - 00:22:58: Yeah.

Noam - 00:22:59: And I admit that when I thought about this problem, I thought it was going to be very similar to Netflix. I didn't realize in biology, there are so many other dimensions, including technical noise, also known as batch effect. And the fact that the number of dimensions is so much higher, et cetera, et cetera, et cetera. So, The problem that we were trying to solve was based on our ability to do the single cell sequencing of the patients. But also to apply AI-based or machine learning-based recommender system algorithms to try and find the right and critical dimensions that make us respond or not respond to drugs. And to leverage compute power that was growing and becoming cheaper and more valuable. And so you asked about the technologies. This was kind of five and a half years ago. This was how we started. And I think today we really doubled down on developing this AI-based capability. To really do this recommendation. And identify the top dimensions, what we call the top-yield features. That make patients respond to the drug or not respond to the drug. And you can look at different variants of this. So this was to the first question that you asked what about like the technologies.

Jon - 00:24:16: That's perfect. That sets the table, right? And now you guys, you stood up a physical lab and you're actually starting to do this. Can you talk a little bit about the physical challenges of getting this done. And then how did you start to educate your potential clients? Did you mention working with Big Pharma? What you described, you described the technology, how, you know, this seems very novel. And it reminds me of what you said earlier, just like Big Pharma had been doing, you know, spending hundreds of billions of dollars doing it in a less efficient way. How did you educate them? And how did you physically get it done and stand up this lab to serve a big pharma partner?

Noam - 00:24:58: Yeah, so there are certain lessons learned that if I were to speak with my younger self, I'm not sure I would have done it again. There are certain stories that are so funny. For example, our lab was set in New York, the original lab. And I remember in one of the first projects, the shipment arrived the same street, but the New Jersey, not New York. Oh no. So this was like something and it was, it's very expensive. And it was because of the shipment made a mistake, not us, but you know, that's, that's how it is. Or in one time there was some problem with the costumes. They just did not release the shipment and it is in dry ice and then eventually, you know, the samples can die. And so I think running a physical lab It is a very complex process. And I think in the beginning, I didn't realize how many different dimensions were going to be into this problem. And that sometimes things can go wrong and you know, everything can go wrong, will go wrong. And the secret to that is really, A, having patience and being able to troubleshoot and just deal with things that, you know, happen and be a firefighter when you have to. And accept it. So there is a very subtle nuance to how do you accept failures without embracing a failure culture so I think this is one area, and I think we really want people to take risks and do things. And sometimes things are not going to be optimal, but we don't want for people to work in an unoptimal way. I think we got it right and it wasn't immediate. So we can talk a lot about like, what does it mean to operate the lab? But this is something that we are not the first ones doing it and we have now amazing lab operations people and we have a large lab in New York. And I think right now it's running like a factory, but in the early days, every project that we did. It was dreadful. I mean, in the sense that we were just every week, we were just praying that we will not mess it up. And you know, that's the secret of like, if you can live with this uncertainty and not die during the process, then maybe you're going to be fine. And that's what happened in the first two years. I think we knew we were taking high risk working with pharma companies. And I had to be very candid with you. It's not that we never messed up, but we never had a, you know, a fuck up. Everything was like manageable. We were able to kind of troubleshoot and eventually figure it out. I think you need to hire people that are very entrepreneurial in spirit and are able to troubleshoot even if they are not founders of the company. And I think we did well. And if we didn't do well, we corrected and we replaced people. But I think we were really investing in hiring the right people. And I think to your second question, which is a more difficult question. How did we educate our pharma partners and what does it mean or what does it take to work with pharma companies? Well, um, inside do you know how you dance with a gorilla?

Jon - 00:28:10: That image in my head is hilarious. Let's say it's a bit imbalanced or a little bit leaning in one direction more than the other.

Noam - 00:28:21: Yeah, so you dance with the gorilla however way the gorilla wants, right? That's the only way. And I think figuring out a way to work with many pharma companies like we had, and we are working now with many of the top 15 pharma companies and like more than 40 partnerships. It's about being able to create a good, you know, balance between doing internal R&D work and working with many gorillas. And I think you need to have experience with your managers in the organization and they're able to, you know, manage accounts and do things that are very high profile, very high intensity and pressure, but being able to manage this to completion, I think we learn from our continued work with pharma companies so much. If you are identifying the right partners, partners is not just the pharma company. It's also the champion within the pharma company, the people that are doing R&D in this organization. They really care about the outcome and they're going to partner with you, they are going to push you, sometimes push you around, but to push you to completion and you're going to learn from them because that's how you develop the product. That's how you develop something that is valuable. Working with follow companies. It's something that requires a very agile and resilient mindset. Because negotiations and sales cycles are prohibitively low. It can take four years to sign a contract.

Jon - 00:29:57:  You said four years?

Noam - 00:29:59: It can take four years. You know, there are some contracts that we signed after four years.

Jon - 00:30:03: Wow.

Noam - 00:30:04: Again, it's about reputation, right? When we started, you know, we were a no-name company. Now it takes less. But yes, there are certain contracts that took us four years from the first discussion to the first pilot and Sometimes you get to the, almost to the finish line. And then you champion, quit the organization because three, four years in today's market, people don't stay usually for seven, eight years. And when your champion leaves, sometimes it's done. It's over. I mean, you need to start over. So you need to have a very agile and resilient mindset and live with not only uncertainty. But with failures. And you need to, how do you inspire people when you just fail and this thing you can't promise will not happen again, it would definitely happen again. So the ability in a company that already raised a lot of money, already has a lot of employees. To not shy away from the complexity, the uncertainty, the failures that happen. This is something that, you know, when you work with large companies you have to figure out a way to doing that. There are some employees that are much more suited to work on internal research and development projects that need the time, they need the patience, they can't stand the lack of stability and constant change, and others, they're just inspired by the change, by the adrenaline rush and all these things, and they work mostly on these projects. And I like this balance. I mean, for me, I probably like the adrenaline rush, but I also like the fact that we are doing very, you know, serious science, and we have some basic science that is being done in the organization, and we work with multiple partners also in academia. So this balance between working for partners that are very demanding, that are very, you know they're demanding in many ways, and you are dependent on them for your growth in some sense, but also having your own backbone and knowing that this is what I need to do for myself, and this is what has to happen for the company to succeed. Balancing the two is very challenging, and I think today we have a very strong and mature leadership and management level that are kind of pushing the company to be able to balance the internal research and development and the external collaborations.

Outro - 00:32:37: That's all for this episode of the Biotech Startups podcast. We hope you enjoyed our discussion with Noam Solomon. Tune into part three of our conversation to learn more about his journey. If you enjoyed this episode, please subscribe, leave us a review and share it with your friends. Thanks for listening. And we look forward to having you join us again on The Biotech Startups Podcast for part four of Genome Story. 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, Excedr. 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.