Notes and tutorials on applied AI, LLM integration and modern web development. No hype, just code that runs.
Field notes on Anthropic prompt caching: how much you actually save, when to enable it, where it breaks. With TypeScript SDK example.
Field notes on agent patterns with the Claude Agent SDK: tool use, MCP, sub-agents, guardrails. What ships to prod, what breaks.
Real differences between AI agents and chatbots in production: when to pick an agent, when to skip it. Concrete client cases.
A hands-on walkthrough of a production-ready RAG pipeline: Postgres with pgvector, embeddings, retrieval and Claude for generation.
A practical guide to plugging an LLM into an existing SaaS: use cases, architecture, TypeScript code, and pitfalls to avoid.
30 free minutes to talk it through, AI or full-stack.