Where Sherlocks AI investigates
Sherlocks AI reads the telemetry your cloud already produces and returns a root cause with the evidence attached. Access is read-only on every platform, and the deployment model is chosen by your compliance posture rather than ours.
sre@prod-bastion ~ $ sherlocks integrations listPLATFORM ACCESS AGENT LOCATIONaws read-only your VPCgcp read-only your VPCazure read-only pending sre@prod-bastion ~ $ sherlocks connect aws --read-only
Start with your cloud
(02 Available, 01 In Progress)- Available now
AWS
Investigate incidents across EC2, RDS, Lambda, S3, CloudWatch, ECS, EKS and SQS. Watson runs read-only inside your VPC, with AWS Bedrock available as a private LLM.
View integration - Available now
Google Cloud
Investigate incidents across GKE, Cloud SQL, Cloud Run, Cloud Functions, Pub/Sub and Cloud Monitoring. Watson runs read-only inside your VPC, and the Fully In-VPC model keeps everything in your project.
View integration - Coming soon
Azure
AKS, Azure SQL, Azure Functions, Service Bus and Azure Monitor, with the same read-only posture. Not live yet: the installer has no AKS path.
Want to see what an investigation actually returns?
An integration page describes what Sherlocks AI can read. The examples section is the other half: real production incidents worked end to end, including the hypotheses that were ruled out. The RCA guides cover the failure classes behind them.
See Sherlocks AI in action
Watch an AI SRE work a real incident from alert to root cause, on your stack, in 30 minutes.