Browse categories
Explore
Fiverr Pro
English
$
USD
I will build a custom rag chatbot to answer based on your documents
Yohan M
About this gig
Your legal team is drowning in documents. I fix that.
Contracts, case files, compliance docs, finding the right answer takes hours. I build RAG chatbots that answer instantly from your documents, with the exact source cited every time.
What I deliver:
- Chat interface connected to your docs (PDF, DOCX, TXT, HTML)
- Source citations every answer references the exact document
- Hybrid search (semantic + keyword) for maximum accuracy
- Multi-document ingestion query your entire document base
- Chat history and context memory
- Clean code + documentation + live demo on your actual documents
Use cases:
- Legal teams querying contracts and case files
- Compliance teams searching regulations and policies
- HR teams navigating handbooks and onboarding docs
Stack: LangChain · Pinecone · OpenAI or Claude · FastAPI · Python
Your data stays yours. System runs on your infrastructure. I never store your documents after delivery. Source code included on every package.
Message me before ordering, free feasibility check, no obligation.
Get to know Yohan M
Yohan M
AI Voice Agent and Automation Engineer
- FromFrance
- Member sinceSep 2024
- Avg. response time1 hour
- Last delivery1 week
Languages
English, French, German, Spanish
I build AI voice agents, automation systems, and custom Python
scrapers that work in production.
Built a full AI cold-calling agent from scratch: Pipecat + Telnyx +
Deepgram + GPT-4o-mini + ElevenLabs, deployed on real calls.
Built custom scrapers for B2B lead-gen on Google Maps, Pages Jaunes,
LinkedIn, and e-commerce sources.
Stack: Vapi, Retell, n8n, Make, Python (Scrapy, Selenium, Playwright),
custom data pipelines.
I debug real problems, not reselling broken no-code templates.
French (native) · English (professional) · German (conversational) ·
Spanish (conversational)
My Portfolio
FAQ
Which document formats do you support?
PDF, TXT, DOCX, HTML, and Markdown. CSV and structured data on request — message me first.
Do I need to host it myself?
Basic delivers a local Python script. Standard and Premium include deployment on your server or a cloud provider (DigitalOcean, Railway, AWS).
Will it hallucinate or make things up?
RAG systems are grounded — the LLM only answers based on your documents. I configure it to cite sources and flag when it doesn't know.
Do I need my own API keys?
Yes. You pay OpenAI or Anthropic directly for API usage. I build and configure the system — you own and run it.
Can it handle large document sets?
Basic handles up to ~50 pages. Standard up to ~500 pages. Premium handles large corpora — message me with your volume.
Is my data secure and confidential?
Completely. Your documents are processed locally or on your own cloud account. I don't store, log, or access your data after delivery. For sensitive legal documents this is non-negotiable and I treat it as such.
What do I receive at delivery?
Full source code, documentation, and a live demo running on your actual documents. You can deploy it yourself or I handle deployment depending on your package.
Can it integrate with our existing tools?
Yes, website widget, Slack, Microsoft Teams, or any platform with an API. Tell me your setup before ordering and I'll confirm compatibility.
