Auto-synced from YouTube. New episodes appear here automatically within minutes of publishing — no manual posting needed.
All Episodes
▶
42:18
Agentic Workflows with Multi-Step Planning
Breaking down how to architect agents that plan across multiple steps without losing coherence — with real implementation patterns from LangGraph and custom frameworks.
▶
38:45
When RAG Works and When It Doesn't
A practical breakdown of RAG failure modes — and which alternative memory patterns actually solve the underlying problem in production systems.
▶
45:02
Tool Calling Strategies for Production LLMs
Why tool-calling is really an interface design problem, and how to structure tools so models use them reliably at scale.
▶
51:30
Evaluating Agents: What Actually Matters
Most eval frameworks measure what's easy to measure. This episode walks through building evals that actually catch the failure modes that matter in production.
▶
36:15
Multi-Agent Systems: Coordination Without Chaos
Orchestration patterns, trust boundaries, and communication protocols — how to make multiple agents work together reliably.