@itanaka2
Isamu TanakaFounder▲5
Founder, agNedia Inc.
18m ago · New project · Agents · Tokyo

Shipped: Kizashi

AI-based failure prediction for network monitoring on Prometheus.

KIZASHI (kizashi, "early sign") brings IT-Service-Intelligence-style adaptive thresholds and predictive alerts to a Prometheus stack. It learns what "normal" looks like for every interface and device, flags deviations, and estimates how long until a degrading metric (optical power, error rate, utilization) crosses a limit, so you can act before the link fails.

It needs no changes to Prometheus. KIZASHI reads metrics through the Prometheus HTTP API and exposes predictions as ordinary metrics on /metrics. Prometheus scrapes them back, so your existing alert rules, Alertmanager routing, and Grafana dashboards keep working.

Kizashi logoKizashi@itanaka/kizashi · github.com/itanaka66/KIZASHI
Kizashi

Comments

No comments yet
Start the conversation.
Log in to join the thread.