01
L1 Market Data Engine
Trading systems require reliable access to real-time market data with minimal latency. Raw exchange feeds are noisy, exchange-specific, and unsuitable for direct consumption by strategies.
Architecture
Exchange WebSocket feeds are ingested and normalized into a common schema. Market metrics such as spread, mid-price, VWAP, and trade velocity are computed in real time and distributed through a FastAPI WebSocket layer.
Challenges
- —Synchronizing trade and quote streams arriving independently
- —Maintaining low-latency calculations without blocking ingestion
- —Designing a reusable schema for future multi-exchange support
Outcome
Built a real-time Level-1 market data platform capable of ingesting, processing, and broadcasting live market information to downstream consumers.

