I will build a private llm or slm chatbot for your company data


Level 2
About this gig
Want ChatGPT style answers from your company documents, without sending your data to a public AI?
I build a private AI assistant that answers from your own files (PDF, Word, Excel, web pages, database) and runs on your server or private cloud. Use a self-hosted open source model (LLM or SLM) or a private Azure OpenAI deployment.
What you get:
- Chatbot that answers from your documents (RAG), with sources
- Self-hosted open source model or private Azure OpenAI
- Login, roles and chat history
- English and Arabic support
- Optional PII masking and audit log for compliance
Why me:
I build AI governance, GRC and compliance platforms for enterprise clients, with semantic search, private AI and PII masking running live in production.
Stack: Python, Node.js, TypeScript, PostgreSQL with pgvector, Docker.
Please message me before ordering with your document types, volume and hosting choice. Server, GPU and AI usage costs are billed to your own account.
Get to know Hardeep Singh
AI Native Engineer, Expert in Replit, Cursor, Claude, Sovereign AI and DevOps
Level 2
- FromIndia
- Member sinceAug 2018
- Avg. response time3 hours
- Last delivery5 days
Languages
Hindi, English, Portuguese, German
My Portfolio
FAQ
Does my data leave my server?
With a self-hosted model, no. With private Azure OpenAI, it stays in your own Azure account and region.
LLM or SLM, which do I need?
An SLM is cheaper and runs on a smaller server, good for focused Q&A. An LLM handles harder questions. I recommend one after seeing your data.
Do I need a GPU server?
For a self-hosted model, usually yes for good speed. Small models can run without a GPU for light use. I suggest the right server size.
Is this fine-tuning?
No. The chatbot answers from your documents. Add new files any time, with no retraining needed.
Who pays for hosting?
You do. Server, GPU and AI costs are billed to your own account.

