Our Story

Back in 2019 we launched our original Carv hardware. Carv 1 was the most technically advanced pressure-sensing insole ever built for skiing. It sat underneath your boot liner and it connected to your phone via a battery pack. Brilliant tech, but not very practical. The insoles affected boot fit, required a special charging cable, and were difficult to switch between boots. 

So, we built Carv 2, launched in October 2024. We dropped the insole and cable, moved a motion sensor into the battery pack and, trained on all of our Carv 1 data, were able to match the accuracy without needing to install anything into your boot. 

Carv 2 was a huge leap forwards, but for customers on a ski trip the friction of waiting for Carv to be delivered was too high. If you hear about Carv on the chairlift, by the time the hardware arrives your trip is over. That got us thinking: how much of Carv’s coaching can we deliver without hardware? As it turns out, a lot more than we expected. 

Which brings us to a new way to experience Carv. It’s the same app but now uses the accelerometer in your phone to measure how you ski and coach you on how to improve. No hardware needed. 

How were we able to do this?

It comes back to something we learned building Carv 2.

Carv 1 measured pressure directly, right under your foot. But alongside those pressure sensors sat an accelerometer and gyroscope, so every session gave us two things at once: what the pressure sensors were reading, and how the boot was moving at that exact moment. With a big enough archive of those sessions, we could see how tightly the two were linked and teach a model to work out what the pressure sensors would have said, from motion alone.

That's what made Carv 2 possible. We no longer needed the insole, because motion had become a reliable proxy for pressure. Simpler hardware, same Ski:IQ.

We've now done exactly the same thing again, one step further out.

Every time someone skied with Carv 1 or Carv 2 while carrying their phone in a pocket, we captured a pair: precise, boot-level measurement of the skiing, and the motion the phone was picking up at the same instant. Same method, new sensor. We could learn how pocket-level movement maps to what the boot measures directly - the g-force through a turn, the rhythm of your edge changes, how smoothly you move between them - and use your phone's motion as a proxy for what we can no longer measure at the boot.

So this isn't a phone guessing at your skiing. It's a phone taught almost 2 billion turns of Carv 1 and Carv 2 data.