You fight with data for a living. So do we… in the open.
We're independent data journalists navigating a world brimming with messy data in the midst of an industry disrupted by AI. We think Data Elixir readers can relate.
That's why we're building open data products and writing dispatches with practical advice and discussion of our strategy and challenges. We're part of a wave of creators building the future by embracing old-school values like transparency and utility instead of fossilized, self-congratulatory myths.
Subscribe and join the conversation: We'll share what we're learning through our projects, and we want to learn from you to make those projects better.
287(g) Watch
A living map and explainers about 287(g) agreements, with auto-generated, multilingual assets along with detailed state and agency profiles.
Mapas Cali
354 map layers of Cali, Colombia, freed from a forgotten server. From beautiful relief map to a census of trees.
Missouri Vehicle Stops
Twenty-six years of data about the most common way people interact with police. Chat about the data or read the latest analysis.
Recovered Factory
You're looking at it. We hired a local street artist to make our logo. We're named for Argentina's "recovered factory" movement.
Questions we're asking
How do we get more out of the stuff we're already building?
We’re empowering people to do more with data. Data journalism is often locked up in rigid articles that can only express so much of much bigger and complex topics.
That’s why we build expressive tools like our Missouri Vehicle Stops MCP for the people who want to go deep. We’re using data to build compelling, free visual assets for social media.
And our dispatches describe practices like our sustainable system for internationalization so you can reproduce the parts that are helpful.
What does rigor mean in a world full of noise?
Old-school journalists wanted certainty: a smoking gun, a single clean finding. Administrative data, at the scale it takes to understand the world, was never going to give them that. The signal was always buried in noise.
If that certainty ever existed, it’s gone. Rigor hasn’t gone anywhere.
The values hold even as the tools change: show your work, check it, estimate the error, let the public test or improve it, say what you see plainly and honestly. None of that vanished — and in an age of misinformation, it matters more.
The path runs through AI, not around it: in the hands of people who go see for themselves, and who use it to hold power to account. That’s still how you tell signal from noise.
We're open to contract work — pipelines, explorers, maps, site translation, charts, MCP servers, and agents for public-interest data. Tell us what you need built.