← Articles
MS

Mikael Sörensen

Author

LLM evaluation engineer at a Copenhagen AI safety company, designing benchmarks for alignment and factuality testing. Built the automated red-teaming framework that runs 10 K adversarial prompts per model release.

aillmtestingsecurity

Articles by Mikael Sörensen(1)

Building a RAG pipeline with LangChain and pgvector

Retrieval-augmented generation transforms a generic LLM into a domain expert by grounding its answers in your own data. This guide walks through building a production-ready RAG pipeline using LangChain, PostgreSQL with pgvector, and the Anthropic Claude API — covering chunking strategies, embedding storage, retrieval tuning, hybrid search, evaluation, and the operational concerns that separate a demo from a deployed system.

raglangchainpgvectorpostgresqlembeddingsaillm

16 July 2026

Mikael Sörensen — ANN Tech — ANN Tech