How do you really know if your latest deployment succeeded?
You can use an AI agent to check!
This is an interesting use case for HolmesGPT that a user told me about on Slack. He is running HolmesGPT in ci/cd to check on new deploys and verify they're healthy. If there is a problem, Holmes investigates and messages his team on Slack.