How I AI

How this PM streamlines 60k-page FDA submissions and saves millions with Claude, Streamlit, and clever AI workflows | Prerna Kaul

245 snips
Jul 14, 2025
Prerna Kaul, a product and platform leader with over 14 years of experience in AI, shares her groundbreaking work at Panasonic Well. She reveals how she developed an AI system that automates 60,000-page FDA submissions, transforming months of work into mere minutes. Kaul explains the importance of transparency in AI cost management and how tools like Claude and Streamlit help non-technical teams. She also addresses ethical considerations in regulated industries, ensuring compliance while maximizing efficiency in life sciences.
Ask episode
AI Snips
Chapters
Books
Transcript
Episode notes
ANECDOTE

Automating FDA Submissions

  • Prerna Kaul automated the creation of a 60,000-page FDA submission that usually took 20 specialists and 4-6 months.
  • This innovation drastically cuts time and cost for getting life-saving vaccines to market.
ADVICE

Set Clear AI Requirements Early

  • Clearly define strict structural and data privacy requirements before building AI solutions in regulated industries.
  • Prioritize choosing AI models that align with safety and compliance from the start.
ANECDOTE

AI Picks Right Models Fast

  • Claude AI rapidly selected the best models for medical entity recognition without deep initial domain input.
  • It saved days of study and effort compared to traditional engineering approaches.
Get the Snipd Podcast app to discover more snips from this episode
Get the app