This is the first post on the new site, so: hello. I'm Amara. I run Sweepr360, which is four people and roughly a year of money. I'll mostly write here about the thinking, the bets, and the runway. Callum handles the parts that are fun to read.
Let me tell you where this started, because it's not a pitch. It's a capstone.
The thing that wouldn't stop getting stuck
My final-year project was a cleaning robot. The brief in my head was simple: don't get stuck under the same chair twice. Every robot I'd lived with eventually wedged itself somewhere and waited to be rescued, and the expensive ones did it just as confidently as the cheap ones.
So I packed the prototype with compute. Bigger brain, the reasoning. And it worked, sort of, and it cost a fortune, and it ran hot, and the battery I needed to feed it made the whole thing heavier, which made it draw more, which meant a bigger battery. You see the loop. Smart, on a moving robot, eats itself.
The measurement
The part that actually changed my mind was boring. I logged how much of the day the robot spends moving versus sitting on its dock. For mine it was a few percent. Call it twenty-three hours a day parked.
I was paying for a brain that's switched off twenty-three hours a day, and dragging it around for the one it isn't.
That's when it flipped. Put the expensive, hot, clever part in the thing that never moves and is always plugged in - the dock. Let the robot be a brilliant, dumb body: good sensors, precise motors, just enough onboard to stay safe and take orders. The dock thinks. The dock learns the home. The dock hands Alpha the right program for the job and Alpha goes and does it.
Cheap body, smart edge
Two reasons this matters, and neither is clever.
Cost. Every gram of intelligence you take off the robot is a gram you don't move, power, cool, or pay for twice. A simpler body is cheaper to build and harder to break - fewer things on the moving part are fewer things that fail on the moving part.
Capability. If the smarts live in the dock and the cloud, then the robot gets better through software, not new hardware. The body you have today does more next month. That's the whole bet in one sentence.
The friends-and-family math
Here's the part I'm not used to saying out loud. This is funded by my savings and by people who know me - friends, family, the kind of money where you've met the person it belonged to. I sold my car for runway. I count it in weeks.
I'm 23. I can barely run payroll. I had a graduate job and I turned it down for this, which my mum is still being very gracious about.
So why bet a company on a hunch about where to put a chip? Because I measured it, and the measurement was honest, and it pointed somewhere nobody was building. That's the most reason I've ever had for anything.
There's an open question underneath all of it, though, and I'm not going to pretend there isn't: can the dock really carry the load I've put on it? Everything rests on that.
I'll show you the working either way.