Founder Stories: Liran Belenzon, Co-Founder and CEO of BenchSci

Welcome to Ask AI, the podcast that brings you insightful interviews and news from the world of Canadian artificial intelligence. This episode is sponsored by Microsoft Canada. Go to microsoft.com/ai and check out their free AI business school to start building intelligence into your solutions today. We're also sponsored by Cinchy, the global leader in data fabric technology. Visit cinchy.com to learn how to eliminate integration and turbocharge your AI transformation.

So Liran, thanks so much for joining us on Ask AI. Could you tell me a bit about yourself and what you do at Benchsci?

Sure. So my name is Liran. I am the CEO and one of the co-founders of Benchsci. And that is what I do at BenchSci.

Could you give us a bit of an elevator pitch for BenchSci?

BenchSci, over the past 6 years, we developed an AI technology that can basically understand the entire history of the medical science like a PhD scientist would, with specific focus on all the evidence that has ever been generated from experiments, which is extremely important because it basically teaches us how biology works.. And by understanding all that evidence at a PhD scientist level, we were able to basically create an evidence map of biology. And that is super, super crucial and important for scientists who are working on leveraging that data to develop life-saving medicine. So what we do with that knowledge graph that we were able to construct is build web applications or software SaaS applications that then scientists at the leading and the largest pharmaceutical companies in the world used throughout the first 7 years of drug discovery, which is anything from an idea to clinical trial, to make sure that we develop drugs faster and they're more successful in clinical trials. And today we work with 13 of the top 20, 8 of the top 10 pharmaceutical companies. We're around 300 people in the company. We raised around $100 million from investors that include Google Ventures, Grady Adventures, which is Google's AI fund, Inovia, F Prime, Will Ventures, Golden Ventures, A4, and more.

Wonderful. Thank you. And the machine learning models that you are building at Benchsci, what area of AI are we talking about? Is it a mix of different types of models? Are we talking about NLP, or could you give me just a summary of the types of models?

Yeah, at a high level, our machine learning focuses both on text, which is very important, but also on images, which are very important. So a lot of the scientific knowledge is described in the text by the scientists of what actually happened in the experiment. But the result of the experiment at the end of the day is a scientific figure, an image, or a graph. So a lot of it is they interpret of what they need for the research, but so much information exists in the image that's actually not being talked about in the paper. Because it's there, but it's not including what they're trying to research or in their findings. And our ability to understand that is very, very powerful because it allows us to understand more and more and more about biology in a way that's just not captured today.

Right. And why is it not captured in the paper? Is there a specific reason? Is this the nature of this type of research?

For sure. It's captured in the image, but not in the paper itself, because it may be not what you tried to research, or you can't what you're trying to show, it's not supported by it, or you don't need to capture it to come to the conclusion or to prove your point or the hypothesis, or just something that you also discovered, but it doesn't mean that that's the point of the paper.

Okay, understood. Thank you. BenchSci is a remote-first company, as I gather. Where are you headquartered within Canada? I'd love to hear more about your remote-first culture, where your offices are located, and how your team, you know, really sticks together in this remote-first environment.

So it's really interesting. So historically, all of us, all the founders are from Toronto. So the company got started in Toronto and before COVID hit, we were around 70 people in the company, the nice office, uh, in Little Italy, Toronto. And in December 2019, we closed our B round. That was led by F Prime. And then we took a massive 27,000 square foot office that can fit up to around 250 people and start the construction. That started in January 2020, and all of us know what happened in March 2020. And obviously we did not know how long it's going to last, and we're already starting renovation. So we just completed it, and we have beautiful 27,000 square foot office in Toronto. And that group of 70 employees grew to 300 during the pandemic. And what happened naturally is a lot of our team members left Toronto because Toronto was really expensive. There was no need to be there because everyone was working from home. And we started hiring a lot of people from pretty much all over. So we have people all over Canada, all over the US, all over the UK. And throughout the pandemic, we saw that being a remote-only company, like back then during the pandemic, works. So we decided to be a remote-first company, which means it's not remote only, but it's remote first, and to facilitate gatherings between our team members to really build that connection and relationships. So people can use our office as much, our offices as much or as little as they choose. And on a quarterly basis, we ask teams and managers to come together and meet and really focus on connecting as humans versus on the work. And once a year we bring the entire company together. So our 27,000 square foot office we kept in Toronto, there are around 130 people, 150 people today in the Toronto area, but only 10% of them use the office on a daily basis, 5 to 10%. And so that's what we have in Toronto for historical reasons. We opened an office in Vancouver because we have around 30 people there and we have we're opening an office in Cambridge, UK soon. And everyone else that works at the company, because we want to make sure we are being equitable in terms of people cannot work from home, need a space sometimes to get out of there, we give everyone, uh, we are giving everyone now WeWork passes so they can work and have a change of scenery where they need to, out of the house, which whatever reason it is. And what was really important as a chip in the culture is actually not doing the hybrid, We don't believe that. It's really important for us to build a culture that is equitable, and we do not believe that by forcing some people to come to the office and others not, um, is the right thing to do because you do create two classes of people or of team members. There is recency bias and other research that shows that people you see more get promoted more, get treated differently. We did not want to create that in our company.

Right, that's great. And sometimes it's hard to update everyone, like if there's water cooler kind of conversations happening and you're not used to having this remote-first culture where everything needs to be online, captured somewhere online. I find that one or the other is what I've seen that works best.

And also what this enabled us to do is provide our team members with more flexibility and more domain responsibility, which is a big framework for us in the company. So because all of us are not at the same place at the same time, work is more asynchronous. So people have more flexibility in terms of the hours they work and the way that fits their lifestyles. We have certain core hours where we expect everyone to be available, but outside of that shape, you work as you see fit. We have this policy called, uh, Roam, like working from home, which allows people to go and work from wherever in the world they want for roughly 3 months. That's very cool. So for those who are Canadians, I spent my winter in Florida, which was really nice. And I know some of our team members did the same thing, or plan to do that in the future. So I think there's a lot of benefit of this structure that allows people to manage their personal life and work life better.

Absolutely. That's wonderful. So let's switch a bit to this massive Series C round that was announced in January, I believe. Could you talk a bit about this most recent round, who led it, the total amount, and plans on what it's going to be used for or what it's been used for so far?

So we raised $50 million or $63 million Canadian. That was our C round. In general, we raised around $100 million US. The reason I'm saying US and Canadian is because Canada, so I never know who the audience is. And the— this round was co-led by TCD, who is a massive private equity from New York. Uh, they were early investors in Netflix, Peloton, and so on. And was also co-led by Innovia, who is a big growth Canadian fund. So they co-led the round with participation from our existing investors, and both of them have been and are amazing and wonderful partners to us. Really help us scale and grow, um, and provide solid guidance and advice. And what we plan to achieve with that money is really expand our software solution into a full platform and bring it to the hands of every scientist in the world that is working on developing life-saving medicine, because we believe and we learned that with this technology, they can do it much more efficiently and faster. Which at the end of the day translates to healthier people and people living longer and getting hope and cure for life-threatening diseases and just impacting the health of everyone around us. And scientists are our heroes, and we want to make sure that we really democratize AI and give them the powerful insights that they need to save lives.

Absolutely. Thank you. So, Liran, the work at Benchsci is very interdisciplinary. How do you deal with that collaboration, that cross-team functionality?

Great question. So most companies are a three-legged stool: product, design, engineering. Bench has a four-legged stool: product, design, engineering, biology, and science. And even though it's one more leg to the stool, it has a lot of complexity and exponential complexity because a lot of companies in our space failed when they design engineering-based solutions for scientists. And a lot of companies in our space fail when they design science-only solutions for scientists. And bringing those two together is something that is hard to do, something that is crucial to do, and is really at the foundation of everything we do here. So constantly marry scientists and engineers and designers and product managers to work together to not only decode the data that we're trying to understand to help scientists, but then also understand the workflows and the interfaces that scientists need and how scientists work to be able to derive value from it. So we have over 30 or 40 PhD scientists working in our companies across different departments, whether it's machine learning, whether it's product management, research, customer success, technical sales across the entire company. We make sure that we bring the scientific perspective to the table to make sure that at the end of the day, everything we do adds value to the scientists because they're the ones that are working on finding cure for life-threatening diseases.

Right. And everything from developing the model to how you present the information is probably—

Yes. And QAing that. Yes. And understanding it and the nuances of it and understanding where machine learning textures So far, and then decision tree maybe is the right way to go or regular code and so on. It's constantly important and better to make sure that at the end of the day, everything we do, and that's what guides us, delivers impact. Because if it doesn't, it doesn't really matter.

Absolutely. And so with these new funds, I'm assuming that there's going to be some hiring happening. What teams are you currently hiring for? And could you talk a bit about the growth plan for your team?

Sure. So we're hiring pretty much for every team in the company, whether it's finance or people team, our sales team, our customer success team, our engineering team, which is doubling and tripling in size, our science team, our product team, our design team, every team in the company, everything. And we're looking for talented individuals who want to do more with their talents and work on something that makes a positive impact in the world with, in a very tangible way that is real, that is making a real impact here and now, and not work on another email automation or marketing or sales software, sales software and so on. Yep. And that's really where the Do More motto is, which is do more with your talent than your typical job, which is help us help scientists, give hope to patients, and solve some of the most challenging problems known, known to humans. And when we hire, it's really important for us not to focus on culture fit, but on value fit. Okay. And investing in diversity, equity, and inclusion and building a company the right way, in a way that we are proud of is extremely important for us. It's so important for us. That's our definition of success. So we define success in the company as success beyond success. It's not only financial success, not only making an impact with your product, but doing it in a way that you are proud of. Something that's really, really important for us. And I think lots of startups are work hard, play hard kind of cultures. So we are a startup that work and do meaningful work and build meaningful relationships and have fun while you do that. So that's really the startup that we are. I would say it's more of a startup for grownups.

Nice, nice summary. And can you talk about these big problems that you're tackling maybe a bit more concretely so that we can all know like what areas of pharmaceuticals or healthcare it affects?

For sure. So really, we really focus on the first 7 years of drug discovery. It's pretty much everything from an idea until it gets to the clinical trial phase. Okay. So that's where we stop. But it goes into clinical trial optimization and so on that we're not playing there right now. The first 7 years of can, how do I cure this disease? To here's a drug that now is to go into clinical trials is where BenchSci is used and where we focus on right now.

And did COVID, the pandemic, that whole experience accelerate things at BenchSci or did it— how did it affect your company specifically?

So I think there's 3 companies as it relates to COVID. One, they were positively impacted by it, but I guess now they're negatively impacted by it, market. There's the ones that haven't been impacted at all that are worse, and the ones that COVID pretty much destroyed those companies. We're in the second group. We capture so much in the discovery stage before it gets into clinical trials. We haven't been impacted yet, but I do think the world is going through an acceleration of digitization. The space that we are at is probably one of the last ones that haven't been re-digitized, and pretty sure we're going to start filling that suit. Understood.

Thank you. So, Liran, thank you so much for joining us on this Ask AI team check-ins. To wrap this up, do you have any calls to action to the people that are listening today?

Yes, we're hiring a lot of talented individuals to join our company. So if you are interested in joining a company that is making a real impact in the world and is impacting the health of individuals, A company that if it doesn't exist, then no one will solve this problem because we're alone in the market and what we're doing is true innovation. If there will be no Benchai, there will be no Benchai and there'll be no solution to the problem that we're solving. And we're doing this with amazing people who work here and doing it the right way. And we're looking for more amazing talent to come and join us across all departments in the company. And if you have an interest in biology, that's great. If you don't, that's also okay. And hook us up, check our website, apply, shoot me an email.

We want to talk to you.

Great.

Thank you so much for joining us. We'll, we'll attach a link to the website and we'll see you on the platform.

Thank you so much. This has been a production of Ask AI, a nonprofit dedicated to advancing awareness and understanding of Canada's world-class artificial intelligence sector. Your host was Carolyn Pelletier. Series producer was Chris McClellan. The series editor was James Fajardo. We'd like to thank our series sponsors Microsoft Canada, producers of the free AI Business School, and Cinchy, the dataware platform that makes integration obsolete. We'd love to hear about your latest research, product launches, or organizational news. To request a team check-in, Visit askai.org.