Something quiet but important just went live at DevLocker.dev, and I want to take a moment to explain why it matters more than it might appear at first glance.
We added Datasets. Seven open-access sports data repositories are now cataloged alongside our 184 APIs and 92 MCP Servers. That number will grow. But before it does, I want to explain why we built this category at all, because the reasoning gets at something fundamental about where sports technology is actually headed.
The difference between data and a dataset is everything.
An API gives you live data on demand. You call it, it answers, you get today’s scores, this quarter’s stats, tonight’s injury report. APIs are the nervous system. They carry signals in real time.
A dataset is something different. It’s a downloadable collection, a structured archive of sports event data, game logs, player statistics, and match records assembled over time and made freely available with no authentication, no licensing fees, no sales call required. It’s the raw material. The foundation. The stuff you train models on.
The rise of AI-powered sports analytics has created enormous and accelerating demand for exactly this kind of high-quality training data. StatsBomb’s event data. The SCORE Network’s curated repositories. These resources give researchers, students, and independent developers access to the same granular data that professional clubs spend six figures to license commercially. That’s not a small thing. That’s democratization of a previously closed market, and it’s happening right now.
Why This Belongs in DevLocker
When we built DevLocker.dev as the discovery layer for sports-tech infrastructure, the mandate was clear: catalog everything a developer, data scientist, or sports organization needs to build AI-connected sports applications. APIs and MCP Servers were the obvious starting point. Datasets were always next, because they represent a distinct and essential layer of the same stack.
Think of it this way. If APIs are the plumbing and MCP Servers are the connective tissue that lets AI assistants talk to live data, then open datasets are the training ground. They are how you build and benchmark the machine learning models that predict match outcomes, evaluate player performance, optimize in-game tactics, and train large language models on sports-specific corpora. They are the foundation of most published sports analytics research. Without them, the AI layer above doesn’t exist.
Leaving datasets out of DevLocker would have been like building a directory of recording studios without mentioning where you get the tracks.
Who this is for.
If you are a researcher or academic working in sports analytics, open datasets are your primary resource. If you are a developer building a predictive model or a sports AI application, these are your training data. If you are a student trying to break into the sports tech industry, they are your sandbox. And if you are a professional organization evaluating what independent developers are building using publicly available data, that data is your competitive intelligence.
Seven datasets are live today. We will keep adding. If you know of one we’ve missed, tell us.
DevLocker.dev. The discovery layer for sports-tech infrastructure. Brought to you by Comunicano.