Doc AI.
Agentic RAG over your document library — with citations, contradictions and a knowledge graph
12 specialist LangChain tools wired through a planner agent
Citation tracer + contradiction finder + hallucination guard
D3 force-directed knowledge graph across the document library
30+ analytics metrics on reading depth, contradictions and overlap
Production agentic-RAG platform with deep document intelligence. Built on LangChain + LangGraph with 12 specialist tools (document_retriever, citation_tracer, contradiction_finder, multi_doc_comparator, quiz_generator, flashcard_generator, summarizer, entity_extractor, knowledge_graph_builder, hallucination_guard, analytics_engine, confidence_scorer). AWS Textract for OCR, Pinecone for vector search, Prisma + PostgreSQL for state, NextAuth for auth. Every answer ships with traceable citations, confidence scores and a hallucination guard. The library view renders an interactive D3 force-directed knowledge graph linking concepts across docs, and the analytics dashboard surfaces 30+ metrics on reading depth, contradiction density, citation overlap and more.
