Yes I have— Cloud Custodian (https://github.com/capitalone/cloud-custodian) relies heavily on its event driven serverless policies to enable compliance across a large AWS fleet— filling in the gaps that are otherwise missing in IAM. Currently there are in the order of magnitude of 1000s of lambdas deployed across 100s of accounts and it definitely does exactly what it needs to do with little maintenance. Monitoring is done with a combination of cloudwatch, datadog and pagerduty so getting alerted on failing or errored invocations is completely built into our workflow.