RAG

By Matthias Petermann / on 11.07.2026

Bullshit in, bullshit out: Why AI needs good foundations

AI can do impressive things, but it cannot compensate for missing structure. In RAG, agentic coding, and AI music workflows, input quality still determines output quality....

By Matthias Petermann / on 21.11.2025

RAG without magic: Auditable knowledge integration for local LLMs

This article explains in a structured and practical way how to build a local, auditable knowledge pipeline for LLMs: Markdown → Chunking → Embeddings → SQLite → Retrieval → Prompt → llama-server. A minimalistic, reproducible, cloud-free, blackbox-free RAG implementation....