Guide
What does it mean for AI to manage your projects?
Lots of companies want to add AI, but the data to add it to isn't in one place. If projects are in one tool, schedules in another and staffing in a spreadsheet, AI has nothing to read.
Written 2026-09-22
For AI to read it, the data has to be on one platform
This is the prerequisite. If project progress, who's assigned where, and how much revenue has come in live in different tools, AI can't connect the three. A question like "who should we add to the project that's behind this month?" only gets an answer when all three are in the same place.
So the order is reversed. The first step isn't adding an AI tool — it's bringing the data together in one place.
Take progress out of people's hands first
If progress is a number typed in by hand, AI can't trust that number either. "About 60%" isn't a counted value, it's a feeling, and a judgment built on a feeling is a feeling.
Sprintingo calculates progress by having the server count the cards in the Done column. The number changes the moment a card moves, and it updates for everyone looking at the same screen without a refresh. It shows overall, development and non-development progress separately for each week.
What AI actually does
- 01It flags projects falling behind firstIt reads completion rates against the weeks remaining, whether weekly goals were met, and blockers written in daily scrums, then picks out and points to projects that are short of target. You know before the weekly report, not after.
- 02It lines up what to do nextIt gathers cards carried over three weeks in a row, cards with no owner and cards with blockers, and shows them in the order to tackle this week. It also tells you when a task should be split.
- 03It proposes staffingLooking at project phases, remaining work, and how many projects each person is on right now, it suggests who to put where.
- 04It shows its evidenceIt doesn't just hand down a verdict. It shows which cards in which week it based its call on, so if you disagree you can check right there.
See AI tool costs on the same screen
Roll out an AI coding tool to your team and a bill arrives next month. Usually, no one knows who's using how much.
Sprintingo shows each employee's Claude Code usage broken down by tokens, API-equivalent cost and registered monthly subscription. It also manages plan registrations, so you don't have to go around asking "who's on which plan." From the Team plan up, you also get a macOS desktop app to switch accounts and see usage limits from the menu bar.
What to sort out before adding AI
- Is progress tallied automatically? — Nothing can be built on numbers typed in by hand.
- Are records piling up a line a day? — If blockers aren't written down, AI doesn't know what's stuck.
- Is staffing in one place? — Staffing plans can't be made without it.
- Are sales and revenue in the same account? — "Which project should get more people?" is a question with money on it.
Frequently asked questions
- Does AI replace managers?
- No. What AI does is find signals in scattered data and surface them first. People decide what to do, and since AI shows its evidence, you can check and decide right there.
- Is our company's data used for AI training?
- Data is separated per company (workspace), and the server checks on every request that you're an approved member of that company. We don't use it for training. If you need on-premise installation, we discuss it as a custom deployment.
- Which plan do AI features start on?
- Every plan gets all features and AI features as is. Every plan has unlimited users; what separates the plans is the number of workspaces and Loopy's monthly AI limit (shared across the company).