Real-time anomaly detection across yourNode.js event pipelines.
Nightjar watches consumer-group health, lagging queues, and the correlating service-side signals. It emits plain-English summaries with the supporting traces attached, then auto-applies the reversible remediations and posts the actionable ones to Slack and PagerDuty for a one-click approval.
Three pillars · one autonomous loop
Detect the stall. Correlate the cause. Apply the reversible fix.
Most AIOps stops at correlation. Nightjar keeps going — proposes a remediation, runs the safe parts automatically, and asks a human for the parts that aren't. Fewer pages, faster answers, and an audit trail end-to-end.
Detection
Real-time detection across Node.js + Redis Streams.
One watcher per stream taps the events BullMQ already emits — consumer-group stalls, PEL creep, lag spikes — alongside the surrounding Node services. No agent to bolt onto the host. No sidecar to babysit.
Root cause
Plain-English root cause.
A spike in one queue against a slow consumer in another is a deployment bug, not two alerts. Nightjar joins the timeline so the on-caller sees the chain — not three independent pages.
Remediation
Safe auto-remediation with Slack + PagerDuty summaries.
Reversible actions apply themselves. Anything that touches data, replays jobs at scale, or rolls back a deploy is held for a human-in-the-loop approval link (PagerDuty) with a Slack summary attached.
Run Nightjar on your fleet
One watcher per stream. One approval channel for your whole fleet.
We're onboarding the first cohort of teams now. Tell us where Redis Streams live in your stack and we'll get you a watcher — usually within a day, on a single stream to start.
14-day trial on a single stream · no agent to install · reversible from day one.