Home / LLM agents & research
AgentQuant
Iterative trading strategy research framework with walk-forward testing
What it is
A Python framework for trading strategy research that iterates on proposals, backtests them and keeps a memory of results in SQLite. It includes holdout evaluation and walk-forward experiments, along with experimental genetic-algorithm and diff-based features. Developers and researchers who want agent-assisted strategy research with safeguards against overfitting are the intended users. Several features are marked experimental, so expect parts of the workflow to change.
At a glance
Research onlyOur rating, based on popularity, maintenance and how ready it is for real use.
| Best for | Professional quants |
|---|---|
| Used for | Strategy research, Backtesting |
| Markets | Multi-market |
| Stack | Python |
| Learning curve | Steep learning curve |
| Practical value | Medium practical value |
| Cost | Free and open source |
| Hardware | No GPU needed |
| Maintenance | Last commit 4 days ago |
GitHub stars, last 30 days
Daily snapshots since 2026-09-28 (up to 30 days): +12 over the period, now 217. Gaps mean no snapshot was taken that day.
Trading-strategy research framework with iterative proposal generation, backtesting, SQLite memory, holdout evaluation, walk-forward experiments, and experimental genetic-algorithm and diff
Similar tools
Vectorized Python backtesting and strategy research toolkit
Research framework for RL-enhanced Avellaneda-Stoikov market making
Python library of quantitative finance building blocks
AI-oriented quantitative investment platform with full ML pipeline
Open-source reinforcement learning platform for quantitative trading
Event-driven backtesting engine for Python trading algorithms