About
distillcore is built and maintained by Fahad Baig.
The ecosystem
Universal document processing library. 7-stage pipeline: extract, classify, structure, chunk, enrich, embed, validate. 5 file formats (PDF, DOCX, HTML, TXT, Markdown), 4 embedding providers, SQLite storage.
Pydantic-AI agent layer. 4-agent sequential pipeline (Triage, Processing, QA, Research) that autonomously configures and validates the distillcore pipeline.