I will build a rag or text to sql ai agent to chat with your data

D
deykumaresh
D
deykumaresh
Kumaresh Dey

Level 2

build a rag or text to sql ai agent to chat with your data
Full Screen
build a rag or text to sql ai agent to chat with your data

About this gig

Want to ask your documents or database questions in plain English and get correct answers? I build RAG and text-to-SQL agents that answer from your own data and show where each answer came from.


What I build:

  • Document Q&A (RAG) over PDFs, Word, Notion, Confluence, Google Drive or your website
  • Text-to-SQL agents for PostgreSQL, MySQL, Snowflake, BigQuery and more
  • Answers with sources or the SQL used, so your team can trust them
  • An evaluation set to measure accuracy before launch
  • Web chat, Slack or Teams bot, or an API your app can call


Why me: in my day job I build exactly this: a RAG data-discovery pipeline and an AI agent that writes SQL and builds BI charts, cutting manual modelling effort by 80%.


Stack: Python, OpenAI / Claude / Gemini or open-source LLMs, LangChain, LangGraph, LlamaIndex, pgvector, Pinecone, OpenSearch, FastAPI.


Tell me what data you have and the questions you want answered, and I'll suggest the right package.

  • Bot type

    • Customer Service & Support
    • News & Information Updates

  • AI engine

    • Gemini
    • Open AI GPT
    • LangChain
    • LLaMA
    • Claude.ai

  • Programming language

    • Python
    • TypeScript
    • React

  • Tools & frameworks

    • Pinecone
    • Hugging Face
    • LangChain

  • Platforms

    • Websites
    • Slack
    • Microsoft Teams

Get to know Kumaresh Dey

Kumaresh Dey

AI Engineer for AI Agents, RAG, MCP and Claude Code, Python Backend

4.9(27)

Level 2

  • FromIndia
  • Member sinceJul 2023
  • Avg. response time1 hour
  • Last delivery2 months

  • Languages

    Hindi, English
AI engineer and Python backend developer with 4+ years of production experience. I build GenAI tooling on an enterprise data platform: RAG pipelines, text-to-SQL, AI agents and MCP integrations on Snowflake and AWS. Before that I built backend systems serving 120k+ users and data pipelines that collected 600k+ records. On Fiverr I set up the right AI workflow and tool stack for your budget, configure Claude Code and MCP servers, build AI agents and RAG chatbots, and deliver reliable web scraping. Send me your use case and I'll tell you plainly what will work.

My Portfolio

Related tags