How SystemFlow is put together
SystemFlow lets teams design, simulate, monitor, and analyze distributed systems. This page covers the core concepts.
The four pieces
Canvas / Topology
Design infrastructure as an executable diagram. Nodes carry real behavior - latency, throughput, failure rates - not just shapes.
Simulation
Run traffic patterns against your topology to see saturation and cascading failures before they reach production.
ML & Anomalies
Adaptive, per-signal baselines and peer comparison flag deviations that a single fixed threshold would miss.
Causal Analysis
When something breaks, the causal engine walks the dependency graph and correlates signals into a likely chain of cause and effect.
How it's built
A browser-first simulation engine, a Node.js backend for auth and project storage, a Python ML service for anomaly and causal analysis, and Go collectors for production metrics ingestion. We don't have a public API or SDK yet - if you need one, tell us what you're building.
Have a question the docs don't answer?
We're four people - email us and a real one will reply.