Sanofi Scales Plai, an AI Quality Intelligence Tool, Across Multiple Manufacturing Sites

Sanofi's Plai, an AI-powered quality intelligence tool, is now used across multiple manufacturing sites globally, according to an ISPE iSpeak write-up of a podcast episode with Miguelina Matthews, Head of Quality Intelligence, Advocacy, and Pharmacopeia Affairs at Sanofi, drawn from her presentation at the 2026 ISPE Facilities of the Future Conference. The write-up says investigations that once took hours or days are now completed significantly faster, that AI-driven pattern recognition improves insight and consistency by reducing variability across sites and individual reviewers, and that automating manual analysis and report drafting lets investigators focus on critical thinking, decision-making and risk assessment. The write-up names trust as the biggest hurdle to bringing AI into a regulated GxP quality environment. Open questions include how to validate AI models, how to explain their outputs to regulators and how to ensure algorithms do not drift over time. It also describes a cultural shift, since quality organizations have long relied on deterministic, rule-based decision-making while AI introduces probabilistic insights. Plai began with a tightly scoped pilot that automated deviation and complaint trending and generated potential root-cause hypotheses. Within months, the team saw faster identification of trends and more targeted investigations. Sanofi embedded change management from the start, involving end users in design, being transparent about limitations and focusing on usability. Governance comes from Sanofi's RAISE framework (Responsible AI at Sanofi Enterprise), which provides guardrails for developing, validating and deploying AI tools. The write-up sums up the approach as starting focused, proving value, governing responsibly, then scaling. The episode also covers two metrics Sanofi uses to understand quality across its network: the Quality Maturity Index (QMI), which measures internal quality performance using indicators such as deviations and corrective and preventive actions, and Quality Risk Exposure (QRE), which adds external signals such as inspection trends and audit outcomes. Matthews predicts that within five years AI-assisted investigations, predictive quality intelligence and automated reporting will be standard practice, and she anticipates deeper integration across quality, manufacturing, supply chain and regulatory systems.