LogSentry, Sensitive Data Detection in Logs

An observability and security-operations system that watches CloudWatch Logs across every service and catches credentials leaked into production logs before they become breaches. Log groups are auto-subscribed (EventBridge + scheduled fallback), streamed to Kinesis, and scanned by a Lambda with 12 regex patterns plus Shannon entropy analysis. Findings land in DynamoDB (with TTL auto-expiry), the scanner emits custom CloudWatch metrics per severity with alarms, and rate-limited SNS alerts fire to Slack/email. A Flask dashboard provides live scanning, filtering, and one-click resolution. Zero-config after deploy, new services are monitored automatically within minutes.