This is a big one! It would be great to have an instance status next to the instance in the portal that would go from green to yellow to red based on info collected from various syslogs and eventually IPMI scraping. When something like MAC flapping happens we could alert the customer by email and change the status in the portal and offer an API endpoint for monitoring. We could now have a "premium" monitoring option that is customer facing and will assist our TAM team with early warning data. This could be a fantastic use for an inference engine.