Greencat-AI
AI Knowledge EngineEngineer & Technical Partner
A retrieval-augmented AI assistant built on a publishing entrepreneur's full coaching corpus. Four ingestion pipelines turn years of Zoom calls, Loom recordings, course lessons, and opening monologues into ~1,800 structured knowledge documents in a continuously-updated vector store, answering student questions in his own voice. The weekly flow is fully automated, so the corpus stays current with zero manual work.
~1,800+ knowledge documents across four content pipelines spanning 2+ years of material; the weekly ingest → index → answer flow fully automated end to end. Bulk question-answering parallelized for a measured 3.3x (projected 5-8x) speedup at no added cost.