Introducing Quadrant: Priority Without the Noise

I've used a lot of task managers. I've also built one. And the problem I keep running into isn't capturing tasks - it's knowing which ones actually matter today.

The Eisenhower matrix isn't a new idea. Urgent vs. important, four boxes, clean mental model. Every productivity book covers it. But almost no tool makes it the primary interface. Most bury it in a "view" option, if they have it at all.

Quadrant is built around that model from the ground up.

What It Is

Quadrant is a priority-first task manager. Every task lives in one of four zones:

• Must Finish - urgent and important. These get done today.
• Schedule - important but not urgent. These get a time.
• Delegate - urgent but not important. These go to someone else.
• Backlog - neither. Figure out where these go else they get cut.

The interface is the matrix. Not a list with a priority field - the actual two-by-two. You drag tasks between zones. The layout forces the question: why is this here?

Why Another Tool

Todo handles day-to-day task capture well. But it doesn't have strong opinions about priority. Quadrant does.

The goal isn't to replace Todo - it's to answer a different question. Todo is what do I need to do. Quadrant is what should I be doing right now, and what should I stop pretending I'll do.

That second question is harder. Most tools avoid it.

What I Learned Building It

Quadrant was built with AI assistance - specifically Claude Code. It wasn't pure vibe-coding; it was closer to a collaborative development exercise. When you have a capable AI writing code, it's easy to let go of the reins entirely. I'd argue that's a mistake.

The model is good, but it doesn't cover every scenario. Security flows get left half-finished. Edge cases get missed. Back-and-forth prompting helps, but human oversight and code review remain essential - not optional. I'll write more about using AI in the SDLC in a future post.

Try It

Quadrant is live at quadrant.withpreet.com. It's early, and I'm still refining the interaction model.

If you try it, I'd love to hear what you think.