A question I got from a coworker: how easy/hard would it be to add a new custom/unsupported data source?

Could you elaborate on what you mean by custom/unsupported data scource? Is it a store where you have metadata stored already?

It's as easy as implementing a new Metadata Ingestion source in Python!

@agreeable-hamburger-38305: we’ve had quite a few contributions in new integrations from people like @square-greece-86505 , @curved-sandwich-81699, @high-hospital-85984 @square-activity-64562 @adamant-pharmacist-61996 and others. They might be able to give you a good sense of how much time adding a new source takes.

It of course depends on the complexity of the source; if it provides good APIs to get the metadata then you could be done in a day or two. If you try pull it from a source that hasn’t been designed for this sort of stuff, then those days could turn in weeks 😅 The Datahub -specific code is quite well documented and there are a ton of examples to follow

Agree @high-hospital-85984 😁 But overall, the guides @big-carpet-38439 mentioning above is super helpful. Covering all the things from where to start, what to be installed to how to test