Tools & Systems
Phase 8 of the quantitative learning path. Once you've built a strategy that passes every validation step, the question becomes: how do you actually deploy, monitor, and maintain it without losing money to operational issues?
8 articles ~1.6 h
Architecture
- Arithmion SDK Overview . The two-layer split between the `apexquant.core` contract tier and the `apexquant.bundled` implementation tier: which layer your code lives in, and the rules it runs under.
- Data Pipelines and Storage . Point-in-time data, the three cleaning operations, and storage formats (Parquet over CSV, why).
- The Arithmion Backtest Engine . Vectorized for research, event-driven for production. The cost model and validation that ship by default.
Deployment
- Research to Production Workflow . The six stages: research → validation → paper → pilot → deploy → monitor. The gates between stages that strategies must pass.
- Live Trading Infrastructure: From Backtest to Real Capital . Broker APIs, position reconciliation, risk checks, and what kills strategies in production.
Operations
- Monitoring and Alerting for Live Trading Strategies . Live Sharpe tracking, IC decay, capacity, fill rate, drawdown alerts. What to watch and when.
- Risk Management Systems for Quant Trading . Position sizing, leverage caps, drawdown controls, stress tests. The five layers of risk.
- Model Monitoring and Retraining: Keeping Alpha Alive . When to retrain (time-based, performance-based, regime-based, data-based). The four retraining triggers.