Quick Start¶

This walks through a first natural-language query, five minutes after cloning the repo.

1. Install and Configure¶

git clone https://github.com/osmanuygar/sqlatte.git
cd sqlatte
pip install -r requirements.txt

Edit config/config.yaml with an LLM key and a database connection — see Installation for the minimal example. Everything else (analytics, scheduler, semantic layer, ops agent, MCP) stays off until you turn it on.

2. Start the Server¶

python run.py

============================================================
☕ SQLatte - Natural Language to SQL
============================================================
Starting server on http://0.0.0.0:8000
============================================================

3. Ask a Question¶

Open http://localhost:8000, pick a table from the schema panel, and ask something like:

Show me the top 10 rows ordered by created_at

SQLatte runs intent detection, generates SQL against the table's schema, validates it, executes it, and returns results with an AI-generated summary. Ask a follow-up ("what about last week?") and it uses conversation memory to modify the previous query instead of starting over.

4. Explore From Here¶

Want to...	Go to
Connect Claude Desktop / Claude Code to your data	MCP Overview
Embed the chat widget in your own site	UI Guide
Improve SQL accuracy with business-friendly table names	Semantic Layer
Schedule a recurring report	Scheduled Queries
Run BigQuery cost/security/performance audits	BigQuery Ops Console
See every config option	Full Config Reference
Deploy with Docker	Docker Deployment