INGOT Brokers
Risk Management & Monitoring System
Real-time market-exposure monitoring with configurable thresholds and automated mitigation, with no deploy required to change the rules.
Live risk thresholds, changed without deploys
The problem
Market risk does not wait for a release schedule. Exposure shifts in seconds, and the thresholds that were prudent this morning can be wrong by the afternoon. A risk system that needs a code deployment to change a limit is a risk system that is always one release behind the market, and in a brokerage that gap is measured in money.
What I built
A real-time risk platform, built on Python with Kafka for the event stream and Redis for fast threshold state, that watches market exposure continuously and acts on it. The defining decision was to make the rules data, not code: operations teams adjust thresholds and mitigation actions through configuration, live, without waiting on a deploy. The engine consumes the same event flow as the trading core, so it sees exposure as it happens rather than on a polling delay.
The result
Risk adapts at the speed the market moves rather than the speed of the release cycle. Operators tune thresholds and automated mitigations on the fly, which turns risk policy into something the people closest to the market can adjust directly. Built on the same event-driven backbone as the trading platform, it reads live exposure off the bus instead of querying for it after the fact.