Jensen: AI‑powered knowledge‑graph platform for client‑specific insights
Jensen builds a living knowledge graph for each client, seeded from a universal concept genome and enriched by document ingestion and AI chat. It stores vectors in Supabase (pgvector) and uses OpenAI embeddings plus Anthropic LLMs to generate and update concepts. The platform is built with Astro on a standalone Node server, making it easy to self‑host and extend. It targets teams that need a dynamic, AI‑driven repository of domain knowledge, offering faster retrieval and contextual chat compared to static document stores.
View on GitHub →nprimm1992-ui/Jensen