ravega.net
Project

Market source published

Trading platform — FastAPI backend, React dashboard, live tick collectors into a time-series store, an options-pricing layer, and a bot engine with pluggable strategies. The backtesting harness is published under Code.

Stack

Python · FastAPI · React + Vite · PostgreSQL · InfluxDB · Redis

Notable engineering

  • Tick collectors stream live market data into a time-series store under a supervised collector manager that restarts a failed feed with exponential backoff.
  • A bot engine runs strategies as pluggable modules against either live ticks or a recorded session, so the same strategy code is used for trading and for evaluation.
  • Option legs are priced with Black-Scholes because the tick feed carries no historical option marks — the model is the only way to evaluate an options strategy on this data.
  • The backtesting harness detects that the feed stamps volume on roughly one percent of prints and refuses to present a volume-weighted band the data cannot support.

Read the source →