Turing Biosystems has developed an interpretable AI that allows the integration of multiple layers of data for clinicians and biopharma companies to optimise treatments and therapies. We sat down with Adam Amara and Rémy Boutonnet, the two co-founders, to talk about what inspired them to found a deep tech startup. 

Capital Enterprise: Let’s start by talking about your backgrounds. How did you decide to found a company together?

Adam Amara: I did a PhD in Manchester in Synthetic and Systems Biology, and at that time I was starting to get interested in biological engineering and computational biology. Notably, the complex systems behind biology, and understanding how components interact with each other – genes, metabolites, proteins, whatever. After finishing my PhD, then I moved into a postdoc in the cancer research branch of the World Health Organisation where we were analysing large cohort studies of cancer patients with complex data sets up to 500,000 patients. I realised there was this pain everywhere of integrating layers of data for clinicians, for scientists, so they could understand the complexities in different layers of data. You have genetics that interact with your metabolism, with your proteins, and you see this in every disease.

Rémy is actually a childhood friend who came to Manchester for a conference on automated reasoning AI. We’d stayed in touch, and so we met there in the Turing Memorial Park in front of the Sackville Building and we were talking about concepts around biology and computation. So I was talking with Remy about some of the techniques I was working on that I was very proud of and Remy pointed out it’s actually quite trivial. They’ve been using it since the 60s in computer science. And that’s how we started to get the idea of merging these two techniques. 

Rémy Boutonnet: Like Adam said, we’ve known each other for a long time, we met in our science club in high school. And then Adam went to Manchester to study biochemistry while I stayed in France. I studied Computer Science in Grenoble and did a PhD in a very niche and fringe area, you could say, because for me the more arcane the better. It’s what we call automated reasoning, which is basically how to give machines and computers the ability to think like humans think. It’s the ability to reproduce human thought in a machine.

Automated reasoning was applied to analyse the behaviour of what we call critical systems – things that ought not to fail, like nuclear reactors, aeroplanes, transportation systems, railways. I mean, if there is any bug in the software controlling a nuclear reactor, a huge number of people could die. And so this is what these automated reasoning techniques are used for, can we foresee these risks at whatever arbitrary point in the future? Can we guarantee that no adverse events will happen? And if it can, what can we do to find a fix? 

During my PhD I went to visit Adam in Manchester. And we thought that those problems around side effects in therapies and treatments can be seen as adverse events that are happening in the human body. The worst thing that can happen is the human body dying, right? So this is an adverse event, and it’s all biological, from the molecular reactions, to the organs, to the entire population, which can be seen as a huge critical system.

And so we thought, what if we reuse these methods of automated reasoning to predict, for example, whether a patient would have an adverse reaction to a treatment. That can happen with gruesome side effects and it’s very bad. So can you think about a treatment or a way to improve it?

Adam: Rémy then went to work for a semiconductor design company. And basically, we got back into contact and talked regularly about those ideas. It was really abstract at the time of our PhDs, and so it took a couple of years for us to come back to all those ideas and exchanges.

Rémy: But it was kind of a crazy idea of bringing things like automated reasoning that had been created for industrial systems, you know, for aerospace and the military, to biology. And so after my PhD, I went to work for one of the three companies in the world that design tools for designing computer chips. The company is called Mentor Graphics and designs these really complex tools that enable you to design computer chips from a very abstract idea down to something that will be manufactured, for example, in Taiwan. I stayed there for three years and that was the final piece, getting inspired from the electronics design world. So now we’re applying that to Turing Biosystems to design treatments, or to improve existing treatments in combination. 

CE: How did you go from the idea stage to building a company and securing investment? Some of these ideas are so new, you must have encountered scepticism. 

Adam: We ended up talking with a lot of pharma companies and clinicians to make sure we were onto something and not just imagining cool tech. We wanted to make sure we developed something that made sense and was really needed. So we eventually found this concept around optimising therapies for pharma companies and clinicians. 

At that stage, we went on the Cancer Tech Accelerator and later the P4 Precision Medicine Accelerator, and this triggered a lot of things, it helped us to really start collaborating with clinicians. We started working in Newcastle in the UK, collaborating with the hospital and university there on immunotherapies for melanoma and lung cancer. This has now expanded to King’s College Hospital, to hospitals in France, a large pharma company – I don’t know if we can say the name – as well as gene therapy companies and so on. So you could say CTA gave us a nice push to get that started. 

CE: What has been the most challenging thing about being a founder?

Adam: For me, it’s been translating or explaining our technology to a wide range of people, notably investors, or people that don’t have such a good background in this space. It’s so easy to talk with clinicians, with our customers, like this biopharma company. They get the problem and they get us, like 15 minutes into the conversation you see their eyes popping. 

With investors, it’s an uphill battle sometimes. You have this contrast between a biopharma person who’s usually very sceptical but they’re speaking to us and looking excited. And then the investor is like, ‘I don’t understand. Is it machine learning?’ And you’re like, ‘No, it’s AI.’ ‘But it’s not machine learning?’ ‘No, it’s different.’ That’s been a big thing to communicate.

Rémy: It’s true, we spent a year or so to finally be able to get a lead investor committed to our pre-seed round. And now that we are developing our company and hiring people, we are facing new challenges and to be quite frank, this is something that nobody talks about with founders. We all assume that it’s just happening in the background, all the nitty gritty details like accounting. And it’s extremely painful. 

We spend a lot of hours on things that are totally non-sexy, that are totally unscientific, and not devoted to our product. It’s accounting and managing money, finance, legal, managing IP, infrastructure management. If I had a message to other founders, it’s that accounting is a beast and legal is another beast, basically.

CE: That’s very interesting what you said, Adam, about how you speak to clinicians and they get it right away. And then you speak to investors, and they don’t understand what you’re talking about.

Rémy: Yeah, that’s a segment of the challenge because we’re at the intersection of a few different areas. When we started, there were many fewer AI companies. Now it’s exploding because everyone is building a startup with an API to GPT3 and saying they have an AI company. So when you do hardcore AI and deep tech, you really need to learn to communicate about what sets you apart. But one of the things that made it easier was getting a big fat check from a pharma company. Many people find that very convincing, like ‘Oh wait, I should listen to them a little more.’

CE: For sure, it’s a specialised type of investor that is going to understand this. My next question is, how were your experiences of Cancer Tech Accelerator and the P4 Precision Medicine Accelerator?

Adam: It’s been a really good experience. They put us in contact with a lot of people in the ecosystem, you know, the London ecosystem. We are both French, before I was in Manchester. And then we’ve got some operations in France, but the people who really helped the company to grow were in London, by putting us in touch with clinicians and investors in the UK.

Rémy: Our two most critical clinical collaborations in the UK were brought about by relationships that we made in P4, either through mentors or fellow founders in the programme. And of course we pitched at two demo days and closed a £889,000 funding round while we were on the programme.

CE: Finally, if you were giving advice to a PhD student who was thinking of spinning out and starting a company, what would you say?

Adam: That’s a very good question. I would say, the first thing is that we forget, as scientists, or as people who love technology, that you have to go out there and talk with real people to understand what the needs are. You need to understand, before it’s built, before it even exists, what people need and how you’re going to sell it to these people. You can build a lot of shit that’s completely useless. So in deep tech, you could have great IP from your lab, but make sure you go talk to as many people as possible and try to understand what they need and how to sell it to them. That’s it.

Rémy: And really think about, what life do you want? Speak to people, as Adam said, and build some things that are meaningful for you and for the people you meet. Try to talk to people who are different from you, with different backgrounds, coming from entirely different fields. Computer scientists don’t talk to drug discovery or biology people at all. And in biology, they were approaching intuitively things we were already doing in computer science. So talk to different people.

And then the other thing is, maybe this is partly a French point of view, it’s not dirty to start a company. If you say that you want to do a company after your PhD, your professors will sometimes look down on you because academia is the glorified path. But making a company is where you can make things happen in the real world for real people, not just writing papers. In the end, most papers are forgotten. So really, you can make your own path as a PhD student, and try to think out of the box beyond what your professors may want you to do.

Thanks to Remy and Adam for speaking to us. For more news from Capital Enterprise, and to find out about new accelerator opportunities for startups, sign up to our monthly entrepreneurs newsletter here