I'm a triathlete, and I built The Split to fix my own training data.
Two things bothered me every season. First, heart rate: a Polar chest or forehead strap is far more accurate than a wrist sensor, but underwater it's not just inconvenient, it's physically impossible — ANT+ and Bluetooth don't travel through water, so a strap under a swim cap has no way to talk to a watch on your wrist. Second, the forgotten T1: press the transition button a few strides too late, and your swim split runs long while your bike split starts short, throwing off pace and heart-rate averages for both.
Neither problem needs a re-recorded workout to fix. Both just need someone to go into the file after the fact and correct it. That's the whole tool.
I spent about a year building, testing, and actually using this on my own training and racing before I considered showing it to anyone else. Once it had quietly become the thing I reached for after every swim and every race, and I'd seen enough other triathletes describe the exact same two problems with no easy option to fix them, it seemed worth making available rather than keeping to myself.
The backend — timestamp alignment, lap recalculation, multisport boundary splitting — is genuinely complicated. I didn't want the interface to feel that way. So rather than adding a toggle for every option I could imagine someone wanting, I spent most of the design effort removing whatever wasn't strictly necessary. Fewer knobs, not more, was the goal on every screen.
Outside of triathlon, I've spent 20+ years in tech as an engineer and a founder. These days that's a co-founder/COO-CTO role at Blackbelts.fi, and coaching other CTOs and CPOs at Britemind. A lot of what I've learned running products and teams found its way into how this tool is built and designed — sport is a pretty good metaphor for the rest of life, it turns out.