

Challenges Slowing AI Adoption in Life Sciences Manufacturing - with Yunke Xiang of Sanofi
Jun 17, 2025
Yunke Xiang, the Global Head of Data Science for Manufacturing at Sanofi, dives into the hurdles facing AI adoption in life sciences. He highlights how fragmented data sources and outdated infrastructure impede progress. Yunke discusses the balance of building versus buying AI solutions and the pivotal role of leadership in nurturing data science. He emphasizes the need for strong governance and collaboration to successfully implement AI. The conversation offers a rare glimpse into the complexities of transforming pharmaceutical manufacturing with cutting-edge technologies.
AI Snips
Chapters
Transcript
Episode notes
Data Fragmentation Blocks AI Progress
- Data fragmentation caused by inconsistent identifiers slows AI adoption in pharma manufacturing.
- Legacy systems and siloed data make tracing manufacturing processes very difficult.
Leverage Generative AI for Data
- Use generative AI tools like those from Snowflake and Databricks to enable natural language interactions with data.
- This helps business users quickly discover data issues and validate KPIs without SQL knowledge.
Build Foundation Before Buying AI
- Prioritize building a strong data and infrastructure foundation before buying AI solutions.
- Without consistent data and good governance, purchased AI tools won't deliver long-term ROI.