Get the latest tech news

Imagining a personal data pipeline

I've been thinking a lot about personal data lately: where it's stored, how to extract it, and what to do with it. Here's where I landed.

The idea of being a "data-driven" human being, making decisions in the name of "optimization," is not appealing to me at all but there are a number of places where I would like to be able to combine data across services without allowing disparate companies to just vacuum it up wholesale. The idea here is that, instead of maintaining a massive universal schema of all the APIs in the world, there is a collection of transformations you can experiment with, modify, and republish, a bit like using and forking a Git repo. This would not change the functionality of the rest of the system, rather it would help bridge the technical gap and provide a simple interface for anyone unwilling or unable to create their own processing and reporting logic.

Get the Android app