I will build an ai rag chatbot for pdfs, tables and diagrams

M
mangesh_5254
M
mangesh_5254
Mangesh W

About this gig

Is the answer buried in a PDF table or diagram?


I build AI RAG chatbots for founders and teams who need answers grounded in their documents. Start with a small text-Q&A prototype, then add visual retrieval, uploads and integration.


My personal research platform covers 499 papers and ~8M processed corpus tokens, with table/diagram retrieval, developer traces and graph-guided search. See the recorded evidence in my gallery.


BASIC $100: 2 text-readable PDFs/20 pages total, local cited Q&A demo, 3 agreed test questions and source code. No OCR, visual retrieval or deployment.

STANDARD: 15 PDFs/150 pages, printed-English OCR, table/figure retrieval, drag-and-drop uploads, history, branded web app and one deployment.

PREMIUM: 30 PDFs/300 pages, Standard plus login, one app/API integration, caching and handover.


Both file/page limits apply; one app and knowledge base. We review samples and acceptance criteria first. GraphRAG builds, complex scans and larger corpora need custom scope. API/database/hosting fees are separate.


Message me with 2 sample PDFs and the question you need answered.

Get to know Mangesh W

Mangesh W

RAG and Voice AI Developer

  • FromIndia
  • Member sinceSep 2024
  • Languages

    English
I build AI applications that turn documents into answers with source citations, including relevant tables and diagrams. My work includes a research assistant over 499 papers, voice-enabled learning tools, and AI lecture-generation workflows. I work with Python, FastAPI, Next.js, React, PostgreSQL, LangGraph, Qdrant, and Neo4j. I can help with RAG chatbots, AI integrations, and full-stack development, with clear scope, working demos, and handover documentation.

My Portfolio