AI Data Analyst Agent

Upload any data, in any format, set a focus, and let the AI generate code, perform the analysis, and create a report—no matter the size or structure.

Comment

Inspiration

I have done hundreds of data analyses in the past. For simple datasets, it often required a lot of boilerplate code to get insights. So I thought, why not use AI to automate this process?

What It Does

  • Upload your data, and the AI will generate a report with key statistics and insights.
  • The report includes visualizations, summaries, and a spoken version of the findings.

How We Built It

Backend

  • Technology: Python backend, connected to the client via WebSocket.
  • AI-Powered Analysis: Uses OpenAI API to generate, execute, and refine the analysis script iteratively.
  • Report Generation: Creates reports in HTML and PDF formats.
  • Voice Summaries: Uses ElevenLabs API to generate a spoken version of the report.

Frontend

  • Technology: Built with Svelte for an interactive and user-friendly experience.
  • Report Display: Provides an intuitive interface to explore results.

Challenges We Ran Into

  • Ensuring the quality and reliability of results, which varies based on data complexity.
  • Pivoting from Streamlit to Svelte for a better UX.

Accomplishments That We're Proud Of

  • Built the entire project in less than two days.
  • Successfully implemented real-time WebSocket communication for streaming AI-generated results.

What We Learned

  • How to use WebSockets to stream AI-generated results and status updates between frontend and backend.
  • Optimizing the AI prompt engineering process to improve analysis quality.

What's Next for AI Data Analyst Agent

  • Local LLM Support: Enhance privacy by running models locally.
  • Interactive Refinement: Allow users to iteratively refine reports.
  • Q&A Mode: Enable interactive querying of datasets for deeper insights.

Built With

  • elevenlabs
  • openai
  • python
  • svelte
  • tailwindcss
Share this project:

Updates