Escanor

Use case

Monitor production in one place

Escanor reads from the tools you already run and puts deployments, incidents, logs and activity side by side. Send your own logs to the ingest endpoint and they sit next to the rest, ready for you or the assistant to search.

Sign up freeRead the monitoring guide

What you see

Deployments
What shipped, where and in what state, across every connected platform.
Learn more
Incidents
Alerts and findings from your monitoring tools, with severity and status.
Learn more
Monitoring
Status, logs and analysis from connected sources and the logs you send.
Learn more
Activity
One timeline of deploys, merges, changes and approvals across your tools and Escanor.
Learn more
Projects and repositories
Everything you run and every repository, with recent activity.
Learn more
Analytics
Trends in usage and activity over time.
Learn more

How it works

  1. Step 1

    Connect your sources

    Connect the monitoring and deployment tools you already use, such as Datadog, Sentry, Grafana, New Relic, PagerDuty, Prometheus, GitHub and Vercel.

    Integrations

  2. Step 2

    Read the logs

    Filter logs by source, level or text, open a row to see what happened around it, and follow new entries live. Source health shows which data is fresh.

    Monitoring

  3. Step 3

    Send your own logs

    Create a log ingest key in Settings and point your shipper at the ingest endpoint. JSON, newline-delimited JSON, plain text and Loki push streams all work.

    Log ingest keys

  4. Step 4

    Ask what changed

    Ask the assistant what changed before errors began. It compares incidents with recent deploys and logs and links the evidence it used.

    The assistant

Send a test log

Create a key under Settings > Developer > Log ingest keys, then send one entry. A 202 response means it was accepted, and the message shows up in Monitoring under the key's source.

curl -fsS https://api.escanor.in/api/v1/observability/ingest \
  -H 'Authorization: Bearer YOUR_LOG_INGEST_TOKEN' \
  -H 'Content-Type: application/json' \
  -d '{"entries":[{"level":"info","message":"Ingest check","service":"example-service"}]}'

Ingest keys can only send logs. They cannot sign in, read logs or call MCP operations, so a leaked key is easy to contain: revoke it and issue a new one. Each request takes up to 1,000 entries.

From monitoring to a fix

When something breaks, the same data feeds the next step. Open the incident, ask the assistant for the evidence, and review a proposed repair. See AI incident response and Autopilot.

Sign up freeWatch the demoBook a call