[eadan]

This is very cool. Relatedly, as a data scientist, I wish companies would expose their APIs through SQL. I've spent a lot of time pulling data into ETL jobs from things like mixpanel, adwords etc., and having a unified interface would make things much simpler.

I'm trying to understand the architecture of Splitgraph. Are all foreign data wrappers controlled directly by you, or can third parties host a database and connect it to Splitgraph in a federation?

[mildbyte]

Currently we control and set up all the FDWs (well, an orchestration layer does it on the fly as the query comes in and routes the query to the correct schema with foreign tables).

You can also run a Splitgraph engine locally and add your own FDWs to it. We have a lot of scaffolding around FDWs to make their instantiation much more simple and even wrote a blog post [0] about adding a custom FDW to Splitgraph.

However, in the future we'll be adding the ability to add your own backend data sources to Splitgraph that it can proxy to (whether as a private dataset on the public Splitgraph instance or as a "data virtualization" layer when you have an in-house Splitgraph deployment).

The cool thing about this is that this can be a single gateway to all your data silos (Snowflake, third-party SaaS, public datasets) that can handle federated query execution, data discovery and access control (e.g. firewalling queries to sensitive columns even if the backend data source doesn't support this level of granularity).

[0] https://www.splitgraph.com/blog/foreign-data-wrappers