Yeah, so for BeanHub, when you push a commit, the system will capture the accounting book into SQL database. Taking from there, I can imagine there are many data prediction model can be applied to the ingested data. I am also thinking to support "CI job" feature (run a task when you push a new commit), not for building software but for your accounting book. So even if it's not supported by the system officially, it's still possible to build your prediction model.
