Show HN: No chair fixed my back, so we built one that won't let you sit still
movably.com
movably.com
Before any of this, I'd mostly solved the pain on my own: a standing-height desk and a Hag Capisco, rotating between sitting, standing, and draping one leg up (we call it the flamingo now). It mostly worked and the pain went away and I could do this for hours. But two annoyances never did. Every transition required a conscious interruption prompted by discomfort. By the time I moved, I'd already been static too long.
Here's the problem as we see it. Standing desks, walking, stretching all work, but only if you break focus to do them. In deep work you don't, so you sink into what we started calling the "static valley": long stretches of stillness you didn't choose — breaking flow to move just costs too much. So we buy the most expensive, most adjustable chair we can — just to survive the valley in comfort. The variable that actually tracks with reduced discomfort in the research is how often you change posture, not how good any one posture is.
So the design problem wasn't "add movement", a lot of chairs have done that. It was: allow gross postural change (full stand, one-leg, lean) while keeping the upper body stable enough that your hands and eyes never leave the work. That decoupling is the whole thing. Mechanically: the seat pan is split into two halves, each on a brushless motor; either half drops away on a prompt, and it senses load so it won't drop while bearing weight. Auto-mode nudges a change every 1–10 min. Happy to get into the interaction, control, and the safety system since a chair has to meet a lot of challenging requirements.
On evidence: last time I showed this on HN (2022) the fair criticism was that I said "studies" and couldn't produce one. We had completed the study, but it was not yet peer reviewed and published. Now I can, University of Waterloo, published in Applied Ergonomics (Noguchi et al., 2023). I'll bound it honestly: n=16, 2-hour exposures. The finding was that every participant who became a "pain developer" while standing did not develop pain in our chair, and task performance held. Small study; I won't inflate it. Since then we've shipped, and we're collecting movement data from early users — happy to share what we're seeing.
One key feedback from Waterloo was that all the options our chair introduces may be a challenge. For example, you can always ignore a prompt, and in practice moving becomes automatic enough that almost nobody asks for longer durations (1 of 50+ users so far). But the software always prompts every 10 minutes. I don't know where the line is. Here's another we still chew on: in the study our early chair had a fixed, non-adjustable backrest and it was compared against a Herman Miller Embody, which has a very adjustable back. Ours did better on the pain outcome. I don't fully know what to make of that, except it makes me increasingly skeptical that "lumbar support" is doing what the industry claims. The other thing in our case, it's also not cheap ($2,499).
Last thing- if I can share some credit: our co-founder Andriy and his team in Ukraine got us out of the gate with really good working prototypes in a matter of months — I still don't fully know how they pulled it off (the same team has since moved into Ukraine's drone industry). Alegre Design in Valencia, multiple-award-winners in seating with mobility and electronics experience, was a lucky find for industrial design. Helbling Technik in Bern brought the appliance- and medical-device rigor to get it to a production-ready state that meets long-life and safety standards. Hardware is hard, and it helps to fall in with the right crowd.