Most AI tools feel like hiring a very fast intern. You give them a task, they do it, you move on. Paperclip is something different — it's closer to building an entire team.
You set up an org chart. A head of operations, a finance assistant, a customer research role. Except every person in that org chart is an AI agent. They have goals, budgets, and responsibilities. They report up and down the hierarchy just like a real company would. When something needs doing, the right agent picks it up — not because you told it to, but because that's its job.
The bit that stands out to us is the cost control. AI agents left to their own devices can quietly spend a lot of money — think of it like leaving every light in the office on, forever. Paperclip solves this by giving each agent a budget. Hit the limit, it pauses and waits. No surprises on your bill at the end of the month.
It's free, open source, and you can run it on your own computer. Nothing goes to a third-party server unless you want it to.
This is early days — but it's the first tool we've seen that thinks about AI workers the way you'd think about a real team. That feels worth watching.
AI agent — An AI that doesn't just answer questions, but takes actions on its own: browsing, writing, sending emails, making decisions.
Org chart — The diagram showing who reports to whom in a company. Paperclip uses this same idea, but for AI agents instead of people.
Open source — The software's blueprints are public. Anyone can inspect it, modify it, or use it for free.
Self-hosted — You run it on your own machine, not on someone else's cloud. Your data stays with you.
Worth asking yourself: if you could hire an AI assistant for each department in your business tomorrow, what would you actually want them to do first?