Alerts, incidents and automation
From the first alert to a reviewed fix
Monitors watch metrics, logs, traces, devices and security events. Related alerts fold into one problem with its likely cause, and Brain and workflows help you investigate and fix it, with a person approving changes to your systems.
- Slack · Teams · PagerDuty · Opsgenie
- Bring your own model
- Approval before change
6 products
Six products from alert to fix
One alerting engine serves every area, so an application error, a device alert and a security detection can land in the same problem.
Monitors and alerting
Monitors watch metrics, logs, traces, real user data, Kubernetes, network devices and security matches, and a custom SQL monitor covers anything else in your workspace. Notification policies route alerts to Slack, Microsoft Teams, PagerDuty, Opsgenie, webhooks or your email gateway, and downtimes hold notifications during planned work.
- Slack · Teams · PagerDuty · Opsgenie
- Webhooks
Problems and cases
Related alerts fold into one problem, with a suspected or established cause and the changes that came just before it. Open a case from a problem to set its priority and keep its timeline and evidence together, and switch the case to incident mode when you need one.
- Change evidence
- Incident mode
SLOs and on-call
Define availability or latency SLOs from traces or logs, and track the error budget and burn rate over 7, 30 or 90 days. Each SLO gets a burn-rate monitor. On-call schedules and escalation policies are Tracix objects; linking them to teams and notification policies takes Brain or the API today.
- Burn-rate monitors
- Schedules and overrides
Anomaly detection
Anomaly jobs learn the normal range of each series, with daily or weekly seasonality when you want it, and flag spikes, sustained shifts and slow drift. Rare-value, population, log-pattern and outlier jobs raise signals for investigation, and the explorer shows what is unusual, where and since when.
- Daily or weekly seasonality
- Log patterns
Brain
Brain answers questions by querying your logs, metrics, traces, devices and security data with your own permissions, on the model you connect: Anthropic, OpenAI or a compatible endpoint. Changes wait for your confirmation by default, and device or security actions go through their usual approval flows.
- Anthropic · OpenAI · compatible
- Your permissions
Workflows and agents
Build workflows on a node canvas with schedule, webhook, chat or alert triggers, or ask Brain to draft one. Steps that change something wait behind an approval, and Auto-Fix proposes code fixes for production errors, opening a pull request only after a person approves it and never merging it.
- Approval steps
- GitHub · GitLab
One platform
Connected to the other areas
Operations is where signals from every area meet.
- ApplicationsTrace errors, faulty deploys and SLO burn raise problems, and Auto-Fix can propose the code fix as a pull request.
- InfrastructureHost, Kubernetes and log monitors use ready-made templates, and Auto-Fix gathers pod events, restarts and deploys as evidence.
- NetworkWhen a collector stops checking in, it becomes the established cause for the device alerts behind it, and config changes appear as change evidence.
- SecuritySecurity events can start workflows, and response steps become proposals a person approves before they run.
Send your first alert where your team works
Create a monitor from a template and route it to Slack, Microsoft Teams, PagerDuty, Opsgenie or a webhook.