Writing on RAG, AI agents and LLM evaluation

Notes from building LLM systems that run in production: what broke, how I measured it, and what fixed it.

More posts

  1. 1 min read

    Running open-weight LLMs on your own GPUs

    When health data cannot leave the building, the model has to come to the data. What self-hosting open-weight LLMs costs, and what it buys you.

    LLM infrastructureRAG
  2. 1 min read

    Human approval steps for LangGraph agents

    An agent that sends reports or changes records should stop and ask first. How I build approval steps, retries and tracing into LangGraph workflows.

    AI agents
  3. 1 min read

    Hybrid search: when BM25 beats embeddings

    Embeddings miss exact codes, product names and section numbers that keyword search finds at once. Why my RAG systems combine BM25 and vector search.

    RAG
  4. 1 min read

    Zero-shot road segmentation with foundation models

    What foundation models like SAM and CLIP got right on road scenes without labelled training data, where they failed, and how we measured it at Continental.

    Computer vision
  5. 1 min read

    Auto-annotating camera and LiDAR data with foundation models

    Labelling camera and LiDAR data is slow and expensive. How foundation-model pre-labels and camera-LiDAR fusion speed it up without lowering quality.

    Computer vision