AI SQL Copilot

Mohammad Ahmadian — AI / Machine Learning Engineer
Target roles: Data Engineer · Analytics Engineer · AI Engineer

Enterprise-Grade Natural Language to SQL and Analytics Assistant

Hiring managers: Demo Website · Dashboard · Architecture · Presentation pack

Positioning

This is not a SQL chatbot.
It is an enterprise analytics assistant: schema-aware NL→SQL, validation/safety gates, explainable queries, and secure read-oriented execution.

Recruiter snapshot

| Item | Detail | |------|--------| | Domain | NL→SQL · Database intelligence · Analytics copilot | | Capabilities | Schema catalog · generate · validate · explain · safe execute | | Serving | /health · /v1/schema · /v1/generate · /v1/explain · /v1/execute | | Integrity | SELECT-only · banlist · LIMIT · sample demo DB honesty |

Evaluation (labeled fixtures)

| Metric | Value | |--------|------:| | Execution accuracy | 1.000 | | Syntax valid rate | 1.000 | | Safety block rate | 1.000 |

Full report: reports/EVALUATION_REPORT.md

Architecture

System architecture (Released)

Quick start

python -m venv .venv
.venv\Scripts\activate
pip install -r requirements.txt

set PYTHONPATH=src
python scripts/bootstrap_demo_db.py
python scripts/run_eval.py
uvicorn sqlcopilot.api.main:app --app-dir src --host 0.0.0.0 --port 8000
# http://localhost:8000/docs

Documentation

docs/00_MASTER_INDEX.md · Doc 13: SQL Generation Pipeline

Limitations (honest)

Sample demo analytics DB — not a customer warehouse
Offline mode uses deterministic schema-grounded fallback (honest mode=fallback)
Optional LLM adapter when LLM_API_KEY is set

Contact

mohammad.ahmadian.dev@gmail.com · github.com/ahmadian-dev · Turkey (GMT+3)

License

MIT — see LICENSE.

Source documentation