In production
Quant Lab
Research infrastructure for systematic trading: data pipelines, backtests, paper trading, broker gateways.
- Linux
- Postgres
- IBKR Gateway
What it does
Quant Lab is the research rig we test trading ideas on: it collects market data, runs backtests with costs that match real order books, repeats them walk-forward to expose overfitting, and runs surviving ideas on paper before a single real order. Most ideas fail there, which is exactly what it is for.
- Data pipelines. Daily and intraday bars, order-book samples and corporate events into Postgres, with gap and freshness checks.
- Honest backtests. Commissions, spreads and slippage measured from real books, not assumed.
- Walk-forward and holdout. Parameter search on one window, judgement on the next, and a final holdout nobody tunes on.
- Paper trading. Live signals, simulated fills and a daily report, side by side with the backtest.
- Broker gateway automation. Keeping an Interactive Brokers gateway logged in, reconciled and watched, with hard risk caps.
Requirements
- Linux, Python 3.10 or newer, Postgres.
- Market data and broker accounts of your own.
Risks
- No strategy in this kit is profitable, and none is promised. Our own research found no edge that survives costs; that result is part of the documentation.
- Trading can lose more than you expect. This is software for research, not financial advice, and nothing here is a recommendation to trade.
Updated: