I will build rag and offline llm chatbots


About this gig
Want a chatbot that answers from YOUR documents, website or database, with no made-up answers?
I build production RAG chatbots. My recent work includes a chatbot for a commercial bank that runs fully offline on a local Llama 3.1 model, and an AI analytics assistant serving KPI insights across 5+ markets for a global beauty brand.
What I build:
- Custom RAG chatbots trained on PDFs, Word, websites, Notion or SQL data
- Private, offline LLM chatbots (Llama, Mistral, DeepSeek via Ollama). Your data never leaves your server
- OpenAI, Claude or Gemini powered assistants
- AI agents and automations with LangChain and LangGraph
- Chatbot REST APIs (FastAPI) and website or WhatsApp integration
- Website crawling, chunking and vector search (ChromaDB, FAISS, Pinecone)
What you get:
- Source-cited answers that reduce hallucinations
- Streamlit or web chat UI, or an API for your app
- Clean code, a setup guide and deployment help (Docker, VPS)
Tools: Python, LangChain, Hugging Face, Ollama, ChromaDB, FastAPI, Streamlit.
Message me with your data type and goal before ordering. I'll suggest the right model and package.
Get to know Awais Ahmed
AI Engineer, RAG Chatbots, LLMs, Computer Vision Data Science
- FromPakistan
- Member sinceMay 2021
- Avg. response time1 hour
Languages
English, Urdu
My Portfolio
FAQ
Is my data safe?
With the offline option, the model and your data stay on your own machine or server. Nothing goes to OpenAI or any third party.
What hardware does a local LLM need?
An 8B model runs on a GPU with 8 GB+ VRAM, or slowly on CPU. I'll recommend a setup for your budget.
