Skip to content

Roadmap

GlowBack is in an alpha stage: the core research loop is usable, but several surfaces are still being hardened before they should be treated as production trading infrastructure.

Current focus

  • Engine-backed research loop: keep the Rust engine, Python bindings, FastAPI gateway, and Streamlit UI aligned around the same executable paths.
  • Documentation and examples: expand runnable examples, clarify alpha boundaries, and keep setup instructions validated by CI smoke tests.
  • Performance baselines: continue collecting benchmark artifacts for maintained hot paths and add regression thresholds once enough history exists.
  • Data quality and reproducibility: strengthen dataset validation, run manifests, replay helpers, and experiment-registry workflows.

Near-term improvements

  • Publish hosted Rust API docs once the crate-level public surface stabilizes.
  • Add more end-to-end examples for built-in strategies, CSV datasets, and API replay.
  • Extend optimization diagnostics beyond best-trial selection to better expose overfit risk.
  • Tighten Docker/API packaging so deployment instructions are covered by the same level of smoke testing as the local quickstart.

Longer-term themes

  • More complete options backtesting on top of gb-options.
  • Safer paper/live trading parity in gb-live, including broker adapters, risk limits, and audit trails.
  • Expanded risk analytics and real-time monitoring.
  • A richer hosted/dashboard experience after the local Streamlit workflow is mature.

See Assumptions and Limitations for the current boundaries that matter before planning production use.