Sentinel AI
Sentinel AI

URL monitoring
with AI, on AWS.

Checks your URLs every 60 seconds. Emails you on downtime. The AI digest reads your last 7 days and tells you exactly what happened.

Your apps, checked
every 60 seconds.

Add a URL and EventBridge fires Lambda on a fixed schedule. Two consecutive failures trigger the down alert — single blips are ignored. When it recovers, you hear about that too.

60s
interval
2×
to alert
Email
alerts

Fault-tolerant
by design.

Each monitor runs in its own error boundary. SES failures fall back to an SQS queue for retry. The Lambda has a dead letter queue. One broken URL cannot affect the rest.

Per-monitor error isolation
SES to SQS fallback
Lambda dead letter queue

AWS.
End to end.

Every component runs on Amazon Web Services in af-south-1. Compute, storage, auth, email. Provisioned with Terraform.

Powered byAWS
LambdaDynamoDBSESCognitoAPI GatewaySQSEventBridge

AI that reads
your infra.

powered by gpt-4o-mini

Monitoring tells you what happened. The AI digest tells you what it means — in plain English, with real numbers. On demand from the dashboard, or automatically every Monday morning in your inbox.

7-day incident digest
Downtime durations, incident counts, and what needs attention — written in prose.
Anomaly detection
Response time spikes flagged statistically before the monitor goes down.
Weekly email, automatically
Every Monday at 7am — no button press, no dashboard visit required.

Alerting that doesn't
cry wolf.

Three layers of noise reduction built in.

Consecutive failures
Waits for 2 failures in a row before alerting. Single blips are ignored.
Flapping detection
4+ state changes in 15 min → one 'unstable' email instead of an up/down flood.
Slowdown warning
MAD-based response time analysis flags degradation before the monitor goes down.

What it looks like.

Click any screenshot to expand it.

Create accountSign in