I will build a rag chatbot over your documents with cited answers


About this gig
Most chatbots confidently answer questions they were never taught. Mine don't. They answer strictly from YOUR documents, cite sources inline, and say "that's not in the knowledge base" when it isn't.
Try a live demo before you order: https://llm-handbook-five.vercel.app
What you get:
- A real coded RAG pipeline (Python/TypeScript, LangChain), not a no-code wrapper
- Inline [1][2] citations linked to your source documents
- Honest refusals with a relevance threshold: no hallucinated answers
- Deployed to your infrastructure (Vercel, Cloudflare, AWS) or mine
- Full source code and clean handover docs
For founders and teams who need support bots, internal knowledge assistants, or documentation Q&A they can trust.
How we work: fully async, in writing, with a work-in-progress draft mid-project. You bring your documents and an OpenAI or Gemini API key. I'll show you how to get one free, you pay the provider directly, and I never take custody of your key.
Premium adds an eval harness measuring groundedness, retrieval hit rate, citation validity and latency, so you can SEE the quality rather than take my word for it.
Get to know Mohammed A.
AI Developer: Chatbots, RAG and LLM Agents
- FromEgypt
- Member sinceAug 2026
Languages
Arabic, English
My Portfolio
FAQ
Can I try your work before ordering?
Yes — the live RAG demo linked in the description (free Gemini key needed), plus a click-through AI product in my portfolio.
Do I need my own API key?
Yes — you bring an OpenAI or Gemini key and pay the provider directly. Gemini has a free tier and I'll walk you through setup in 5 minutes.
What document formats work?
PDF, Word, Markdown, HTML, plain text, or website pages. Something unusual? Message me before ordering.
Can it go on my existing website or app?
Yes — delivered as a standalone web UI, an embeddable widget, or an API endpoint your app calls.
Do you do video calls?
I work async in writing — you get documented decisions, faster turnaround, and a mid-project draft to review. Ask me anything in chat, anytime.
Who owns the code?
You do. Full source code, deployment, and handover docs are included in every package.

