Project Deep Dive
AI-Powered Data Insights Tool
A full-stack analytics tool that turns natural language into Pandas workflows and interactive data exploration.
Overview
This project combines a React analytics interface with an AI-assisted backend that turns natural-language questions into executable data analysis steps. The focus was faster analyst feedback loops without removing visibility into the underlying transformations.
Key Outcomes
Improved data query speed by about 90%.
Added automated data health reporting and forecasting to support quicker exploratory analysis.
Combined AI-assisted analysis with a full-stack interface instead of limiting the project to a backend prototype.
Architecture Diagram
Natural-language analytics requests move from the React UI through prompt translation, Pandas execution, and visualization output.
Architecture
A React frontend captures dataset context and natural-language prompts for analysis and visualization.
FastAPI handles orchestration, query translation, execution boundaries, and response shaping for the UI.
LLM-assisted prompt translation generates Pandas-oriented analysis steps to accelerate exploratory workflows.
Forecasting and chart modules sit alongside the core analysis path so users can move from query to trend inspection in one interface.
Stack
- Python
- FastAPI
- React
- Pandas
- OpenAI API
- Scikit-learn
- Chart.js