Skip to content

README.md

Latest commit

 

History

History
79 lines (69 loc) · 3.61 KB

README.md

File metadata and controls

79 lines (69 loc) · 3.61 KB
Raw

SQL-Mind 🧠

A production-ready Text-to-SQL SaaS platform that converts plain English questions into optimized PostgreSQL queries using LLaMA 3.3 AI. Features JWT authentication, encrypted database connections, AST-based SQL validation, 3-signal confidence scoring, and real-time query execution with a modern dark dashboard.

🚀 Features

  • 🔐 JWT Authentication — Secure register/login system
  • 🗄️ Database Connection Management — Connect any PostgreSQL database securely
  • 🤖 AI-Powered SQL Generation — LLaMA 3.3 (via Groq API) converts English to SQL
  • ✅ SQL Validation — AST-based parser blocks all non-SELECT statements
  • 📊 3-Signal Confidence Scoring — HIGH / MEDIUM / LOW confidence badge
  • 📝 Query History — Every query saved with results and confidence score
  • 🎨 Dark Premium UI — Built with React.js + Tailwind CSS

🛠️ Tech Stack

Layer Technology
Backend Python 3.11, FastAPI, SQLAlchemy 2.0
Database PostgreSQL 16, Alembic (migrations)
AI Engine Groq API, LLaMA 3.3 70B Versatile
SQL Validation sqlglot (AST parsing)
Frontend React.js, Tailwind CSS, Zustand
Authentication JWT, Bcrypt, Fernet encryption
Architecture Repository Pattern, Service Layer

🏗️ How It Works

┌─────────────────────────────────────────┐
│           USER (Browser)                │
│   Types: "Show top 5 customers"         │
└──────────────┬──────────────────────────┘
               │ HTTP Request
               ▼
┌─────────────────────────────────────────┐
│         REACT FRONTEND                  │
│   (localhost:5173)                      │
│   Shows UI, sends request to backend    │
└──────────────┬──────────────────────────┘
               │ API Call
               ▼
┌─────────────────────────────────────────┐
│         FASTAPI BACKEND                 │
│   (localhost:8000)                      │
│                                         │
│   Step 1: Verify JWT token              │
│   Step 2: Read database schema          │
│   Step 3: Build prompt for AI           │
│   Step 4: Call Groq AI (LLaMA 3.3)      │
│   Step 5: Validate SQL (SELECT only)    │
│   Step 6: Run SQL on user's database    │
│   Step 7: Calculate confidence score    │
│   Step 8: Save to history               │
│   Step 9: Return results to frontend    │
└──────────────┬──────────────────────────┘
               │
       ┌───────┴────────┐
       ▼                ▼
┌──────────┐    ┌──────────────┐
│ Groq AI  │    │  PostgreSQL  │
│ LLaMA 3.3│    │  Database    │
│ Generates│    │  Stores:     │
│ the SQL  │    │  - Users     │
└──────────┘    │  - Connections│
                │  - History   │
                └──────────────┘

SQL-Mind 🧠

Tests