Natural Language SQL

Imagine if you could query your knowledge base with SQL.

select n.*, s.*, ... from notion n where n.person = "jonny apples" join slack s on n.person = s.person

Context

Started listening to Lenny’s Podcast to get a sense for how great PMs think about product. A question I particularly like is about the speakers favorite interview questions. My recent favorite is from from Laura Modi (Bobbie) - “Teach me something”

My workflow is to listen to podcasts on overcast and then share it to notion if there is something noteworthy. I will add the note in the comments. I recently had a use case of compiling all such interview questions and was looking for a way of extracting this information from notion.

ASIDE: Notion’s API has no LIST method for any of its entities, to get documents, you need to use the search API

Looked into pre-built ways of getting the data.

• airbyte: no comment support
• notion (built-in zip): no comment support
• metano: they use the airbyte container, so same limitations apply

The next step of this project would be to ingest comments, probably by forking the airbyte notion connector. But this was too much of a lift given a passing curiosity so I’ve settled on recording the progress here.

Really, what I want at the end of the day is a natural language SQL interface over my notion so I can do something like the below:

select '%interview questions%' from notes where title LIKE '%lenny podcast%'