Sanofi needs to speed up and de-risk the development, scale-up, troubleshooting, and control of drug manufacturing processes that traditionally require thousands of manual solubility measurements, extensive real-world experiments, and coordination across fragmented data, equipment, teams, and external manufacturing partners.
Reported impact
Sanofi reports faster process development, fewer unnecessary physical experiments, better process understanding and optimization before large-scale manufacturing, faster troubleshooting for external partner production, improved productivity and product quality goals, and quicker readiness for manufacturing and control.
Sanofi built an in-house AI-enabled manufacturing R&D stack that combines robotics, predictive models, digital twins, and generative AI across process development and scale-up. In early CMC work, the Solvify solubility prediction tool ingests public and Sanofi-owned chemical data to model likely solubility behavior and prioritize which experiments matter most. Those predictions guide Solutron, a robotics platform that executes repetitive solubility experiments with precision and reproducibility, producing fresh experimental data that feeds back into the models as a self-sharpening loop. Sanofi’s Kinetic AI platform then helps scientists interpret reaction results, rate-limiting steps, and impurity risks, while LLMs simplify scientist-to-robot interaction, experiment design, and navigation of the broader data ecosystem. In manufacturing scale-up, process digital twins simulate equipment and chemical or biological processes, test operating conditions, and troubleshoot partner production issues without waiting for physical experiments. Generative AI tools and agents also assemble reports and gather data to support faster decisions, monitoring, and knowledge sharing across Sanofi’s manufacturing network.
Public and Sanofi-owned chemical data is modeled in Solvify to choose starting conditions; Solutron runs the prioritized solubility experiments and returns new experimental data to update the models. Kinetic AI and LLM-based tools help scientists interpret results and plan next experiments, while process digital twins use process and partner-site data to simulate manufacturing conditions, test fixes, and support reporting, monitoring, and scale-up decisions.
"Doing physical experiments to replicate the issue and identify solutions could have taken months. But digital approaches enabled the scientist to find a fix in just a few days."
"Generative AI tools are used to create reports, enabling better and faster decision-making. ML tools are used to identify optimal production conditions, increasing productivity."
"Large Language Models (LLMs) are used to simplify the connectivity between the scientists and the robots, design new experiments and navigate the data ecosystem more easily."
"Process digital twins – models of equipment and biological or chemical processes across manufacturing steps – enable teams to simulate the production of a molecule."
"Sanofi’s Kinetic AI platform illuminates results from chemical reactions – helping to understand complex reaction mechanisms, what’s rate-limiting, what types of impurities could be formed, and more."
"Solutron has taken over this type of experimental work, accelerating process development with precision and reproducibility – methodically generating data that feeds back into AI models, enabling continuous learning."
"the 'Solvify' solubility prediction tool integrates a vast array of public and Sanofi-owned chemical data, creating models and narrowing the focus, determining what physical experiments Solutron conducts."
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
In-house Sample Jul 28, 2026 Novartis Novartis R&D needs to turn large volumes of scientific, clinical, operational, and real-world data into faster, higher-quality decisions across drug discovery and development, reducing cycle times, surfacing safety… less than 10 percent of drug candidates historically reach the market AI models · AI-enhanced analytics · Generative AI In-house Sample Apr 13, 2026 Elevance Health Elevance Health needed to find fraud, waste, and abuse in very large volumes of claims, payment, contract, and medical-record data so it could detect billing errors, suspicious coding patterns, and improper payments… across an annualized volume of nearly 23,000 reviews In-house Sample Feb 24, 2026 Sanofi Sanofi needed a faster, more objective way to decide which R&D programs to advance, accelerate, or stop while weighing costs, clinical trial enrollment timelines, unmet need, competition, and each program's probability… more than one billion data points across Sanofi Aily Labs · Plai 