Fireside Chat, Interview
Bruno Gagliardo, Global Head of Data & AI @ Sanofi : Molecule to Market, Driving AI Transformation
- Sanofi intends to apply AI across the entire drug discovery lifecycle, aiming to accelerate the current 10 to 14-year timeline from discovery to patient access, with specific use cases covering research, development, clinical trials, manufacturing, supply chain, and commercial operations.
- Immediate applications include utilizing multi-agent capabilities via Snowflake to process millions of datasets and reducing clinical dossier compilation time from 19 weeks to five weeks, while manufacturing and supply chain use cases target root cause analysis for batch failures.
- To support widespread adoption, the company plans to upskill employees in data and AI literacy, hire personnel with relevant technical skills, and launch a concierge tool for administrative tasks, all while pivoting to an agile architectural mindset due to monthly technology landscape changes.
- Governance and value realization are prioritized through the RAISE framework for security and ethics compliance with the upcoming AI Act, the implementation of LLM-based guardrails to prevent model drift, and a "hundred million BOI impact" threshold for prioritizing use cases.
- The organization is currently building a generic platform to deploy multiple agents, with a strategic plan to evolve from a "build-by-borrow" approach to a "buy" approach as the market matures, ensuring quality, security, and cost remain primary factors for production.
- A significant transformation effort regarding data foundations, governance frameworks, and operating models continues, acknowledging that while industry maturity has steeply increased, some customers must still establish robust data bases to mitigate "garbage in, garbage out" risks associated with Generative AI.