The Data Exchange with Ben Lorica The Truth About Agents in Production
19 snips
Dec 31, 2025 
Join Samuel Colvin, founder of Pydantic, Aparna Dhinakaran from Arize AI, Adam Jones at Anthropic, and Jerry Liu of LlamaIndex in a fascinating conversation about Agentic AI. They explore impressive agent architectures, the advantages and challenges of multi-agent systems, and innovative memory and state management strategies. Aparna emphasizes the importance of observability with evals, while the group shares thoughts on bridging technical and non-technical users through no-code solutions. They also discuss future capabilities and realistic expectations for agent technology.
AI Snips
Chapters
Transcript
Episode notes
Type Safety Boosts Coding Agents
- Coding agents have outperformed expectations this year and benefit greatly from strong type safety.
- Samuel Colvin argues type-safe frameworks (like TypeScript-style) make coding agents far more reliable in production.
Use Evals Continuously
- Use evals early and continuously when building agents to discover failure modes.
- Aparna Dhinakaran says teams that use evals are the only ones she sees with working agents.
Start With User Problems
- Start with real user problems and design agents as product solutions, not as AI-first experiments.
- Adam Jones urges product teams to ask users what problems to solve and then apply AI.