See https://github.com/kubernetes-sigs/cluster-api-provider-openstack for Matt's design.
In this Spike, we want to evaluate the work to be done and whether it can live in an EPIC that can be achieved within one cycle.

Outcome: a list of tasks to do and their time estimation, documented in an EPIC.

Spike: https://docs.google.com/document/d/1tnlS9hxowcpT-srAECKNWabi8kqyeD6tgons2VJC50Y/edit