Postgres + pgvector + OpenAI embeddings = productive semantic search in 200 lines of PHP. The tutorial I wish I had read.
Semantic search is no longer a luxury. Postgres + pgvector + OpenAI embeddings = production-ready in an afternoon.
pgvector covers 95% of needs up to 10M vectors without adding a new service.
Embed with text-embedding-3-small, store with HNSW index, query with cosine distance, optionally fuse with lexical search via RRF.
Auth, streaming, prompt caching, tool use, error handling — everything to cleanly wire Claude into a production PHP project.
How to give an LLM access to your data without fine-tuning. The architecture behind 80% of serious chatbots in 2026, decoded.
From Copilot to Claude, AI assistants radically changed how I code in 2026. Concrete feedback after 18 months of daily production use.
I design and ship tailor-made applications. 9 years of experience. Let's talk over a coffee (or a call).
Start a project