All workConfidential · Fintech · Distributed systems · 2022
Real-time arbitrage detection under adversarial conditions
High-frequency pricing across multiple exchanges, under hard latency constraints
TypeScriptNode.jsPythonSolidityRedisWebSockets
The problem
Detecting price differences across multiple exchanges fast enough to act on them, in an environment where every defect is financially exploitable by someone actively looking for it.
The approach
- 01Built real-time arbitrage detection ingesting high-frequency pricing data from multiple exchanges under strict latency budgets.
- 02Designed transaction-control and rate-limiting mechanisms for a high-volume exchange platform.
- 03Worked across TypeScript, Node.js, Python and Solidity with security treated as a correctness property rather than a checklist.
The outcome
Latency-bounded detection running against live market data.
Transaction controls that held under volume and under attack.