My engineering background led me to run tech stacks in global finance. Building systems that could handle large data sets and assess risk was my daily work for years. I found that same care for detail again in my talk with Frédéric Cogny, co-CEO and CPO of Nova In Silico.
Frédéric has an unusual career background. A graduate of École Polytechnique, with a focus on financial math, he first worked on trading floors. From London to Tokyo and Singapore, he worked as a quant, in product structuring, and then in trading. He also led teams of analysts and devs as a Managing Director at Standard Chartered Bank.
Today, he uses that scaling know-how in a very different field: human biology. The core challenge has changed. It is no longer about tracking the swings of a financial product, but about predicting how a drug candidate may act in the human body.
Software engineering meets research
The company grew out of scientific work in Lyon. Frédéric first backed the original project, Novadiscovery, in 2010. He joined the operational team in 2016 as CTO.
Structuring scientific models
When he arrived, he found a group rich in special skills: biologists and drug experts, but no shared engineering process. Researchers were working alone, often with Fortran or Matlab.
He set out to build one common framework and turn those models into an internal simulation library. A reliable software architecture is essential to scale scientific research. This is also a merging challenge that we often see in custom work with tech and data leaders.
The launch of Nova In Silico
To support that goal and make use of the existing talent, the company took a new form in June 2024. Frédéric Cogny, now CPO, and Grégoire Boutonnet, then COO, bought the company and launched Nova In Silico.
That fresh start marked a phase of streamlining and growth. The company became profitable within six months. Today, it has about sixty people and is hiring hybrid profiles, including engineer-biologists from BIM (bioinfo and modeling) programs at schools such as CentraleSupélec and Polytechnique.
A mix of consulting and software
Nova In Silico is built on a hybrid offer. The company provides scientific consulting services and access to its owned software platform, Jinkō. These two sides support each other. The work of modelers in the field always enriches the tool, while the platform boosts the team's analysis power.
Understanding biology through QSP modeling
The consulting team helps labs design clinical trials with math modeling.
Living systems need a deeper method than pure stats. In biology, and especially for new treatment types such as bispecific antibodies in oncology, historical data is often too limited to train strong predictive models on its own. Regulators, such as the FDA and EMA, also want to understand a molecule's mode of action before they approve test plans.
This is where quantitative systems pharmacology, or QSP modeling, comes in. It turns the biology of a disease and the drug-body link into math terms. Each model part can be traced back to published papers, which keeps the work transparent.
Digital twins and virtual patients
To run these models, the company built Jinkō, a clinical trial simulation platform. The software creates groups of digital twins, or virtual patients, whose biological traits are modeled in real time.
Labs use these tools before advanced clinical stages to compare dose plans across thousands of virtual patients. This is the core of digital twin clinical trials. Simulation is also used during due diligence, both scientific and financial, to estimate the value of a molecule before it is bought. Today, Nova In Silico works with clients in the United States, Europe, and Japan, and the US market has been quick to use these in silico clinical trials.
An open setup for augmented scientists
Artificial intelligence is changing how software platforms work with experts.
AI-assisted models
At Nova In Silico, AI does not replace the researcher. AI-assisted model building cuts the time needed to get a working prototype. Artificial intelligence is used to speed up literature review and pull equations from medical research. Scientists still control every hypothesis, and they act as augmented experts who can cut months of work down to a few days.
The platform is also open to outside systems so research silos can break down. Jinkō offers an API and SDK that work with the Model Context Protocol (MCP). This setup lets AI agents connect to simulations as a simple tool, while humans keep the final read of the results.
This kind of advanced AI also calls for strict software design to protect data sovereignty, which is a top need in the clinical field. The platform is also available through research spaces such as Claude Science.
Predicting clinical trial outcomes
The value of this mechanistic clinical trial simulation shows up in its predictive power. On November 8, 2026, at the American Heart Association congress in Chicago, Nova In Silico shared strong results.
During an oral session, the company unveiled Lp(a)HORIZON, an ASCVD heart disease model that simulated Novartis's Pelacarsen treatment on more than 20,000 virtual patients. The in silico study predicted the trial outcome before the results were public. That direction was later confirmed in early September 2026, when the lab said the trial would stop.
Day to day, Nova In Silico's experts work closely with the labs' own modelers, much like an extension of their research teams. That is similar to how we work in software engineering, where adding tech and data experts helps organizations move faster with the right skills at the right time.
By combining applied math with deep knowledge of human biology, Nova In Silico offers a strong technical base for researchers who want to improve clinical work.
COMPANY PROFILE
→ Name : Nova In Silico
→ Leadership : Frédéric Cogny (co-CEO and CPO) and Grégoire Boutonnet (co-CEO and COO).
→ Industry : Pharmaceutical R&D / Clinical Pharmacology / QSP / Clinical trial simulation / Software publishing
→ Why Nova In Silico Stands Out : It blends expert services with predictive tech. The company uses quantitative systems pharmacology to explain biology, showing that models in healthcare must stay transparent if they are going to be used.
Sources
- European Journal of Preventive Cardiology| EAPC | Oxford Academic - Prospective in silico prediction of the Lp(a)HORIZON trial outcome before the announcement of the main results
- Madame LeFigaro - A new quantum leap
- LinkedIn - Jinkō is available at Claude Science
- Medscape - Will we soon be able to conduct clinical trials on virtual patients?
