GraphPulse

GraphPulse is a web app (with optional lightweight desktop agent) that monitors real operational networks—microservices call graphs, data pipelines, IAM relationships, or dependency graphs—and flags structural changes that correlate with incidents. Instead of only tracking CPU/latency, it tracks network-theory signals: centrality spikes, community fragmentation, new bridge nodes, and cascading-failure risk scores. You connect sources like Kubernetes service maps, OpenTelemetry traces, AWS/GCP IAM graphs, or CI/CD dependency manifests. The app builds a time-series of graph snapshots, then alerts when the topology drifts beyond learned baselines. It includes a “blast radius” simulator: remove a node/edge and estimate downstream impact. This is a combination traditional + AI app: traditional graph analytics plus ML-based anomaly detection and natural-language incident summaries. Realistically, it sells best as a focused SRE/Platform tool, not a generic network-theory playground.