manish@os:~$whoami

System Online--:--:-- --
DA

DataWeave AI

v1.4

RAG Platform

StatusBETAEnvproductionUpdated5 months ago

About the Project

A full-stack RAG application that lets users chat with PDFs using LangChain agents, vector search, and OpenRouter LLMs. Answers are grounded in the source documents with inline citations, table reasoning, and authenticated, persistent chat history.

Problem

Long PDFs hide answers. Plain LLM chat hallucinates and can't point to where an answer came from.

Why I built it

Built to make document Q&A trustworthy: retrieve the right chunks, reason over tables, and always cite the source.

Key Features

  • Chat with any PDF
  • Cited responses
  • LangChain agents
  • Vector search
  • Table reasoning
  • Multi-document chat
ClientServiceDataExternal

Key Outcomes

  • Cited answers users can verify against the original document
  • Table-aware reasoning, not just plain-text extraction
  • Authenticated chat history persisted across sessions

Repository Signals

Mock · GitHub not connected
Stars

96

Forks

14

Commits

412

Contributors

2

Watchers

11

Open Issues

5

Languages

  • TypeScript 71%
  • Python 21%
  • CSS 8%

Recent Activity

pushed 3 months ago

Tech Stack

Next.jsLangChainOpenRouterMongoDBVector Search