Introducing Efficient Labs: tools built at the speed of a conversation

For most of my career, the idea was the easy part. I've spent more than two decades in product and marketing, figuring out what teams should build and how to talk about it. The one thing I could never do was build it myself. I'm not an engineer.

So every demo, every interactive explainer, every little internal calculator I dreamed up followed the same path. I'd write up the requirements, pitch it around, and try to earn enough buy-in to get it on the roadmap. If it cleared that bar, it joined the backlog to be prioritized against everything else the company needed, and waited for an engineering team that already had a full plate. None of that is broken. It's how a responsible company spends scarce engineering time. But it also meant a lot of good ideas quietly died in the queue, and the ones that survived could take quarters to show up.

I joined Efficient Computer because the Electron E1 general-purpose processor, and the Fabric architecture underneath it, are genuinely new, and something that new deserves more than a slide. What I didn't expect was how quickly I could tell that story myself. Working with Claude, the gap between "I have an idea" and "the team can use this tomorrow" basically vanished.

The Board Viewer tool came first. I wanted something an engineer could open in a browser and click around, instead of digging through PDFs and datasheets to find a single component or header. I described it, Claude built it, and I started sharing it. On a Zoom call, someone watching me poke at it said, "it'd be great if it could tell you how to wire up a peripheral." I pasted the transcript into Claude. About 10 minutes later the Wire-it-up feature was live: pick what you want to connect, and it tells you which pins to use, how to set the switches, and hands you copy-paste C code. A comment someone made in passing was a working feature before the call had even wrapped up.

Then it just kept happening. The Pin Mapper tool makes assigning pins a simple drag-and-drop exercise, and warns you the second two choices collide. The EVK Getting Started Guide walks you from unboxing the board to running your first program. The Energy Profiler tool graphs power across every rail in real time and tells you what a specific chunk of code actually costs, in millijoules, so "which part of my program is the expensive one" stops being a guess. The Lifetime Modeler tool took the math our CTO built and gave it an interface worthy of it. His numbers, our front end.

None of these took weeks. Every version was a conversation instead of a ticket, and honestly that changed the work itself. When a revision costs you an afternoon instead of a sprint, you stop settling for the first thing that works. You keep going until it actually feels right.

Here's the thing. We build the most energy-efficient general-purpose processors you can get today, and our effcc Compiler is what makes that possible, getting everything out of the hardware so customers can drop in the code they already run in production, with no heavy rewrites. That same instinct shows up in how we spend everything else, too: our capital, our resources, and our engineers' time. AI isn't always right, and we don't pretend it is. But used well, these tools let us deliver the kind of experience our customers deserve, and turn ideas into real products they actually love.

That's why we built Efficient Labs: a place to get tools in front of customers fast, listen, and iterate. We already meet engineers inside the toolchains they use every day. They deserve tools that are simple, clear, and easy to live with, and Labs is how we hold up our end.

Come look around at labs.efficient.computer. If something's missing, or you just have a question, email labs@efficient.computer. There's a good chance it becomes the next thing we build.