Topline Meet The 27-Year-Old Building $1.5B AI Employees (Decagon CEO)
46 snips
Sep 7, 2025 
Jesse Zhang, CEO of Decagon, leads a $1.5B startup that’s revolutionizing customer service with AI agents. He delves into the future of AI in enterprise solutions and the challenges of automation. The conversation covers unique pricing strategies influenced by customer needs and the evolving competitive landscape of AI. Jesse also discusses the delicate integration of AI in businesses, highlighting how it can enhance productivity while creating new job opportunities. His insights provide a fresh perspective on the dynamic world of AI startups.
AI Snips
Chapters
Transcript
Episode notes
Don’t Deploy AI Where No Fallback Exists
- Avoid deploying AI where you need near-perfect deterministic outputs if the model's error rate can't meet that bar.
- Prioritize use cases where human oversight can correct mistakes or where probabilistic answers are acceptable.
Test With Conversation Unit Tests
- Simulate thousands of conversations as unit tests before production to build confidence in AI behavior.
- Add supervisor models to enforce rules and catch unsafe outputs during live interactions.
Wrappers, Not Models, Capture Value
- Wrappers that solve business problems capture pricing power even if they rely on foundational models.
- Depth of orchestration and tooling around models creates defensibility, not just the model itself.