## Overview It's operationally challenging and complex to provide a Shared Runner Services model hosted on public cloud (AWS, Google Cloud, Azure) virtual machines. There are no native features in GitLab to simplify maintenance or provide clear operational guidance to [Ingrid](https://about.gitlab.com/handbook/product/personas/#ingrid-infrastructure-operator). In addition, Ingrid has no easy way to optimize which machine types to use for CI job execution across the dimensions of cost vs. performance. # Business outcomes - [ ] Improve customer satisfaction and retention for accounts (GitLab SaaS or Self-managed) that manage build agents (a Runner Fleet) at scale.