For humans.
These are the projects where a person is the user: someone sitting at a keyboard, with AI working for them. Some are businesses and some are infrastructure, and each one started with a conversation one of us couldn't put down.
The machine does the heavy lifting. The judgment, the taste, and the reason it exists stay human.
Obviously Not Patents #
Find the strategic concepts and distinctive engineering in your code.
Most developers don't realize how much of their code is worth protecting. ObviouslyNot scans a codebase and surfaces its strategic concepts and distinctive engineering, each with a link to the source, a score and a plain-language explanation, so you can see what might be worth protecting as IP. It's open source and runs on your own machine. When you decide to explore a patent, you bring your attorney a clear write-up instead of starting from a blank page.
Local Alaska News #
Community newsrooms powered by AI and government transcripts.
Thousands of government meetings are livestreamed every week, and almost nobody covers them. Alaska News turns those transcripts into draft articles with AI, and then the people who actually live there review, edit and publish them together. The expensive part is automated and the editorial calls stay human, which is what makes local news economically viable. It started in Alaska and now runs in other communities through Communities News.
Election Voting Monitor #
Measuring what AI tells voters about elections.
More and more people ask AI who's winning, what a candidate believes, or whether their vote matters. Voting Monitor asks consumer AI products about real races every day, saves every answer, and compares what they say with prediction markets and the polls. It doesn't forecast, and it doesn't endorse anyone. It keeps a daily record of what AI is telling voters, so drift, bias and disagreement with reality stay visible.
Locally Distributed AI Compute #
Turn idle computers into an AI compute fleet.
Local AI Herd lets the computers on your network, Mac, Linux or Windows, run AI models together. It finds the machines on its own, sends each job to the one best placed to handle it, and keeps going if one drops off. Everything runs on hardware you already own, with no cloud and no API keys. Your machines already know how to work together. They just needed a shepherd.
Belto AI Karaoke #
AI writes the song. You give it the voice.
AI can write a song in seconds, but it can't feel one. On Belto, the AI track is the rough draft and a person steps up to sing it, and the crowd votes the best human take onto the stage. A song isn't really finished when it's generated. It's finished when someone sings it like they mean it.
de-amplify social media #
Making the feed's "stop" actually stop.
Engagement feeds are in court for addicting kids, and the argument keeps getting stuck on whether addiction counts as a diagnosis. de-amplify argues for a simpler fix it calls control integrity: if a platform offers a brake, the brake has to work and it has to hold. The site has a thirty-second test you can run on your own feed, the legal background, and a two-part proposal that regulates the loop instead of the speech.
Sunset #
Not everything stays running, and that is the point. Building something, learning what it had to teach, and then shutting it down cleanly is its own kind of win. These are finished, not forgotten.
- Keeni.Space 2019–2022
Standard operating procedure software, born out of rocket launch operations. It turned static documents into procedures a team could actually run together, step by step. The problem was real. The market was slower than the runway.
Build with us #
We're building tools that put AI to work for people. If that sounds interesting, we'd love to talk. See the Hiring page, or our open source work on GitHub.