🧠 NL2SQL AI Assistant

A powerful AI-driven platform that converts natural language (NL) into SQL queries, executes them on structured data sources (JSON, CSV, MySQL, SQLite), and provides rich visualizations and automated reports/presentations. The system features voice input, an autonomous AI agent, and a clean Next.js frontend for seamless interaction.

---

📸 Screenshots

Chat View	AI Chat Flow	Report Generation

Home Page	Discover Datasets	Workspace

User Profile	Normalized Tables - 1	Normalized Tables - 2

---

🚀 Features

• 🗣️ Natural Language to SQL Conversion
  Convert plain English questions into accurate SQL queries using LLMs and agents.

• 📊 Auto Visualizations
  Generate bar, line, pie, or tabular visualizations based on SQL query results.

• 🧾 PDF Report / Presentation Generation
  Export query results and visualizations as downloadable PDFs or slide decks.

• 🎙️ Voice Input Support
  Interact with the assistant via voice commands for hands-free querying.

• 🧠 AI Agent Framework
  Handles query generation, error retries, validation, and follow-up suggestions.

• 💾 Multi-format Data Support
  Load data from:

  • JSON
  • CSV
  • SQLite
  • MySQL

• 🛠️ Built-in Tools

  • Load DB: Upload and manage structured datasets.
  • Generate Query: Automatically generate SQL from natural language.
  • Load Query: View, run, and edit previously generated queries.
  • Retry on Error: Automatic query rewriting if an error occurs.

---

🧑‍💻 Tech Stack

• Frontend: Next.js
• Backend: FastAPI & Flask
• Database Support: MySQL, SQLite, CSV, JSON
• AI Stack: OpenAI / Gemini (customizable), LangChain (Agentic execution)
• Visualization: Plotly, Matplotlib, Seaborn
• PDF/PPT Export: Python-PPTX
• Voice Recognition: Whisper

---

🧠 AI Agent Flow

1. Receive NL input (text/voice)
2. Generate SQL using an LLM
3. Validate SQL – retry on errors
4. Run query on selected DB
5. Generate visualization based on result type
6. Export (optional) as PDF/PPT
7. Suggest follow-up questions