Scaling AI to over 50,000 Users: Lessons from the Field – with Harsha Gurulingappa, Merck
Click on the button to load the content from Spotify.
“Where we are moving towards is more transactional. It’s not just about consuming. It’s not about question answering. It’s all about also to be able to perform tasks.”
In this episode, Carsten Bange and co-host Florian Bigelmaier talk with Harsha Gurulingappa from Merck about what it really takes to scale AI from pilots to enterprise-wide impact – reaching up to 50,000 employees with an internal AI companion. Harsha explains why phased rollouts outperform “big-bang” launches and breaks down the non-negotiables for scaling responsibly: governance and traceability, data privacy guardrails, security, and legal/regional compliance. The conversation also explores how Merck’s AI strategy is evolving from classic RAG-style knowledge retrieval toward transactional and agentic AI that can execute tasks in systems of record.
Harsha Gurulingappa on LinkedIn.
Carsten Bange on LinkedIn.
BARC on LinkedIn.
Stay up to date with our newsletter.
