Find out what changed, without setting thresholds

Vigilon checks your apps, endpoints and jobs every five minutes against their own history, and opens an insight that says what changed, where and since when. Serious ones reach your inbox. Included with Team and Pro.

What you get

Six checks, running all the time.

Error rate spikes

Opens when an app fails at least twice as often as it usually does at this time, and turns critical past 10% errors.

Slower endpoints

Opens when an endpoint's slowest requests move into a slower band and it is clearly slower on average.

Error surges

Opens when one known error suddenly happens many times more often than its usual hourly rate.

Missed job runs

Opens when a scheduled job stops running on its schedule, and turns critical after two missed runs. It watches jobs scheduled at least ten minutes apart, once three of their runs are on record.

Failing job streaks

Opens after three failed runs in a row, turns critical at five or after a day of failing, and names the release if the failures began with one.

Setup gaps

A quiet note when your telemetry is missing something the other checks need, such as an app version or route names.

How it works

Nothing to configure, no thresholds to set.

  1. 01

    Send your telemetry

    The same requests, errors and job runs your other features use. There is nothing extra to install.

  2. 02

    Vigilon learns normal

    Each check compares against its own history: the same hours in past weeks, the previous day, or an error's own rate. Error rate and latency checks arm after a day and get sharper after a week.

  3. 03

    An insight opens

    It says what changed, where, since when and against which baseline, and closes itself once things are back to normal.

  4. 04

    Serious ones are emailed

    Critical insights email your team once. Warnings and notes stay in the dashboard, and your coding agent can read them over MCP.

Questions

Anomaly detection questions.

What is anomaly detection in application monitoring?

Anomaly detection watches your app's own numbers and flags a change from what is normal for it, instead of waiting for a fixed threshold you set. Vigilon runs six checks on your apps, endpoints and jobs and opens an insight when one finds a change.

Do I have to set thresholds?

No. Each check compares against its own history, such as the same hours in previous weeks or the previous 24 hours, so there is nothing to configure.

How often does Vigilon check?

Every five minutes. Error rate and latency checks arm after a day of data and get sharper after a week.

Which insights are emailed?

Critical ones, once, to the recipients your admins chose. Warnings and setup notes stay in the dashboard. Admins can turn insight emails off for the workspace.

Which plans include automatic insights?

Team and Pro. Starter includes error tracking, tracing, endpoint and job monitoring, and the MCP server, without insights.

Can my coding agent read insights?

Yes, on Team and Pro. The MCP server's list_insights and get_insight tools return each insight with its severity, its readings and the baseline it was compared against.

Get started

Start free for two months.

Pick a plan, add the SDK, and Vigilon starts watching your app. Starter is free for your first two months, with no card needed.