Hands-on Workshop: Building Production AI Agents

What you’ll learn

Spend a full day building, evaluating, and shipping production-grade AI agents alongside practitioners who do this for a living. We’ll move fast, write a lot of code, and ship working systems by the end of the day.

Morning — Foundations

• Agent architectures: ReAct, plan-and-execute, multi-agent
• Tool-use design and function calling in practice
• Memory: short-term, episodic, vector — when to use what
• Evaluation: trace-based testing and offline benchmarks

Afternoon — Production

• Sandboxing and permission models for agents that touch real systems
• Cost control, fallbacks, and graceful degradation
• Observability: tracing, replay, debugging non-deterministic systems
• Shipping to users — rollout strategies, kill switches, feedback loops

Who should attend

Engineers and tech leads who’ve shipped at least one AI-assisted feature and now want to graduate to autonomous workflows. Comfort with TypeScript or Python required.

What’s included

• Full-day instruction with two practitioners
• Working code repository with starter templates
• Lunch and coffee on site
• Six-week follow-up Slack access for participants
• Recording of all sessions

Featured speakers

• Alex Johnson
• Maria Chen