, 2025-04-16 13:56:00
In this interview, CellVoyant founder and CEO Rafael Carazo Salas explores how AI, live-cell imaging, and stem cell technologies are converging to transform cell therapy development. He shares insights on the challenges of scaling living therapies, the role of predictive AI in clinical trials, and how spinning a biotech company out of academia can drive meaningful impact in the real world.
Let’s start with CellVoyant – what inspired you to launch the company, and how did your background in academia shape the vision behind it?
The inspiration for CellVoyant really stemmed from a moment of clarity – what I like to call our eureka moment. As a Professor at the University of Bristol, my research focused on combining cutting-edge imaging technologies with computational data analysis to understand and design cells with enhanced functions.
Over time, it became increasingly clear that if we could train AI models to analyze and interpret cellular behavior, we could radically transform how we develop cell-based therapies.
In essence, CellVoyant is the bridge between academic insight and practical impact. Our vision is to make it possible to monitor and to predict cell behaviour – to see the future of cells – at scale and use that foresight to transform how we develop therapies.
CellVoyant combines AI, live cell imaging, and stem cell tech. Can you walk us through how these elements come together in your platform to accelerate cell therapy development?
It starts with live-cell imaging. We capture continuous, high-resolution data on how stem cells grow, move, and undergo changes like cell differentiation in real-time. This gives us a dynamic view of cellular behavior that static snapshots can’t provide.
Then, we apply AI, leveraging powerful proprietary AI models. Our models analyze these rich datasets to identify subtle patterns that might be invisible to the human eye. Crucially, we use AI to predict how a cell will behave in the future so that we can select and design better cells from the very beginning.
There’s been a lot of talk about the commercial viability of cell therapies. From your perspective, what are the key factors that will determine whether these therapies can scale and become accessible to more patients?
These therapies hold immense promise, but moving from lab-scale development to large-scale manufacturing is incredibly complex. Cells are living, dynamic entities, and even the smallest environmental changes – temperature, pH, nutrient availability – can impact their behavior and viability. That makes consistent, high-quality production extremely difficult.
Batch failures are unfortunately common, and when you’re working with expensive cell culture media and other materials, the costs add up quickly. Each failed batch not only represents a direct financial loss but also sets back timelines and puts pressure on already limited development and manufacturing capacity.
The key to making cell therapies truly scalable and accessible lies in using predictive, data-driven approaches. If we can use AI to understand better and control cell behavior early on, we can improve consistency, reduce waste, and build smarter, more resilient manufacturing systems.
It’s about shifting the paradigm from reactive to proactive, end-to-end optimized processes because only then can we truly bring these therapies to more patients faster and at a significantly lower cost.
Image Credit: Shutterstock.com
One of the major challenges in cell therapy is predicting how cells will behave over time. How is AI helping CellVoyant tackle this, and what kind of impact do you think this could have on the broader biotech field?
Traditionally, much of cell therapy development has relied on static snapshots of cells or time-consuming, trial-and-error approaches. But cells are dynamic, different from one another, and ever-changing.
By combining live-cell imaging with CellVoyant’s proprietary data and AI, we can develop models that continuously monitor, interpret and understand how cells’ behavior evolves in real-time. These models can predict cells’ future, what type they’re likely to become, how functional they will be, and whether they could be suitable for therapy.
Cells can carry out complex, sophisticated functions that no conventional drug could ever deliver. By making cell therapies predictively, scalable, and accessible, CellVoyant’s vision is to enable the cell therapy industry to deliver its full-fledged promise and, quite simply, revolutionize biotech and medicine as we know them.
Clinical trial failures are notoriously expensive. How can AI help reduce this risk earlier in the development pipeline, particularly when it comes to safety assessments?
AI has the power to transform how we run clinical trials and support scientists every step of the way. Much like how Google Maps provides navigation and enables the correct travel route in real-time, AI, alongside real-time imaging, provides actionable insights that can help improve and accelerate the development of medicines throughout the clinical trial pipeline.
By step-changing our capacity to assess and prognosticate cell quality, we can understand how best to manipulate stem cells efficiently and cost-effectively while minimizing the risk of failed clinical trials and significantly reducing scientists’ manual workflow.
When it comes to safety assessments, AI will be able to combine drug and patient data, to predict which patients are most likely to benefit from a cell therapy treatment in a clinical trial and help predict side effects.
Overall, we see AI as a tool to de-risk, de-cost, and potentially accelerate clinical trials, ensuring they can be run as efficiently and effectively as possible.
When you think about the future of cell therapies for diseases like cancer, diabetes, or neurodegenerative conditions—what excites you most about the potential impact?
What excites me most is that cell therapies have the potential to support millions of people by developing treatments for diseases that are currently untreatable, such as dementia.
We are already witnessing clinical success with CAR-T therapies for blood cancer, such as in the UK, with the National Institute for Health and Care Excellence recently approving lisocabtagene maraleucel for lymphoma patients.
We are also seeing groundbreaking developments for other conditions. For example, patients with wet age-related macular degeneration received a treatment derived from stell cells and reported regaining enough vision to be able to read.
What’s important now is ensuring these therapies are affordable and accessible for all. They are often prohibitively expensive, putting them out of reach for the very people who need them most.
What are some of the biggest barriers—technical, regulatory, or otherwise—that still need to be overcome before cell therapies become mainstream?
Some key barriers to overcome include meeting stringent regulations and the significant time required to develop, manufacture, test, and approve cell therapy candidates.
As previously mentioned, AI could help completely redefine how we tackle these barriers by helping scientists work faster, at lower cost, and with unprecedented predictive power.
For instance, it will help streamline and iteratively improve R&D through optimized automation and development workflows.
When it comes to regulation, it will allow cell therapy developers to ensure compliance and traceability by tracking and organizing vast amounts of data. It will also reduce the time needed for regulatory submissions, testing, and approval, which can be incredibly time-consuming.
Founding a company spun out of a university is a unique journey. What was that process like for you, and what do you think universities can do to better support innovation and commercial spinouts?
I spun CellVoyant out of the University of Bristol in 2021, and it was amazing that I could take inspiration from an idea that my research group had been working on and explore its big-picture commercial applications.
The experience of spinning out of a university is so valuable because it bridges the transition between researching and exploring a scientific idea and seeing how it can work in the real world. The University bet on the ideas I wanted to explore and gave me the freedom to explore and develop them fully.
Universities in the UK do a good job of supporting innovation. They have the connections and mechanisms in place to help a startup access investors, talent, and more. In particular, Bristol has a thriving biotech ecosystem and punches above its weight in scientific talent.
As a whole, UK universities ask for a significant stake from the outset in a spinout, which contrasts with the approach of the US. This can be a challenge and make spinouts seem much less appealing to investors. Though there are many things to improve on that front, and it’s an intense ongoing conversation, things are gradually and increasingly changing for the better.
As someone deeply immersed in both cutting-edge science and entrepreneurship, what trends in AI-biotech are you most closely watching right now?
AI-first companies are revolutionizing the drug discovery landscape, and the approval of the first AI-driven drug design (AIDD) is something that many of us eagerly await to demonstrate how this new way of developing medicines can ultimately bring tangible results.
The trend I am most focused on is how the financial and investment landscapes evolve in response to the current geopolitical tectonic shifts we are witnessing in 2025. These changes will inevitably bring opportunities to redefine where talent is directed, where investments are made, and, ultimately, how best to steer the industry going forward.
And finally, what advice would you give to scientists or researchers looking to transition their ideas out of the lab and into a startup environment like you did with CellVoyant?
Take the leap and do it! It is a big responsibility and requires a lot of energy, but it is thrilling to see your ideas come to life.
I’d urge you to learn from your peers and founders of similar companies, who, in my experience, have been generous in sharing their successes and failures.
While it may seem like the unknown, remember that you are well-placed to meet the challenge and can apply your wealth of experience as an academic, just in a different way.
Where can readers find more information?
About Rafael E. Carazo Salas
Rafael is the founder and CEO of CellVoyant, an AI-first biotechnology company with a mission to accelerate cell therapy development with the power of AI.
He is also Professor and Chair of Biomedical Sciences at the School of Cellular and Molecular Medicine in the University of Bristol, formerly at the University of Cambridge.