Introduction to Agents

Lead author, with Antonio Gulli, Shubham Saboo, Michael Zimmermann, and Vladimir Vuskovic, on Kaggle, November 2025

Day 1 of the Kaggle × Google 5-Day AI Agents Intensive, which reached 1.5M+ learners. It is written for developers, architects, and product leaders moving from proof-of-concept to production-grade agents:

• Taxonomy: Level 0 (core reasoning) → Level 4 (self-evolving systems)
• Architecture: Model (“brain”), Tools (“hands”), Orchestration (“nervous system”)
• Agent Ops: LM-as-judge quality, metrics-driven go/no-go, OpenTelemetry traces, human feedback
• Security: the utility vs. risk trade-off, defense in depth, agents as a new class of principal (SPIFFE identity), least-privilege policies
• Governance: a central gateway/control plane instead of agent sprawl
• Interop: agents with humans, agents, and money (A2A, AP2)

I also joined the Day 1 livestream as a guest.

The Full Series

1. Introduction to Agents (podcast)
2. Agent Tools & Interoperability with MCP (podcast)
3. Context Engineering: Sessions & Memory (podcast)
4. Agent Quality (podcast)
5. Prototype to Production (podcast)

Why This Matters

Building a prototype is easy. Making it secure, high-quality, and reliable is the hard part. Success doesn’t come from the first prompt. It comes from engineering rigor across the whole system: tool contracts, error handling, context management, and comprehensive evaluation.

Read the whitepaper: Kaggle