6
8 Comments

Show IH: I built a local AI companion for the Oura Ring it learns your body, runs entirely on your desktop, and never sends data to a server

Built Vitra: a local AI companion for your Oura Ring (Mac/Win). It learns your body on-device and gives a plain read — push, rest, or hold. Finds patterns like alcohol → HRV −18%. No cloud, no account, one-time price. Feedback welcome from anyone who tracks HRV.

submitted this link to Icon for group Show IH
Show IH
on May 13, 2026
  1. 2

    Been using Vitra for a couple weeks now — solid companion to my Oura Ring. Having my readiness right there on my work machine and home desktop means I'm not constantly pulling out my phone to check, which sounds minor but adds up. The extra insights have been the real surprise; picking up patterns I'd never have spotted on my own. Nicely done.

    1. 1

      Thanks really glad it's earned a spot on your desktop. The patterns being the surprise is the thing I most hoped for; that's exactly why it learns your own baselines over weeks rather than comparing you to an average. Appreciate you taking the time to say so.

  2. 1

    Quick bit of backstory since this is Show IH:

    I built Vitra for myself first. I'd worn an Oura for years and the thing that always bugged me is that every score is graded against a population average: "your HRV is good for people like you." But I don't care about people like me. I care about me. A 62ms HRV might be a great day for one person and a warning sign for another.

    So the core bet is: run a small model on your own machine that learns your personal baselines and patterns over weeks, instead of comparing you to a curve. After ~14 days it starts calling things specific to you: e.g. for me, alcohol drops my next-day HRV ~18%, and a room under 19°C buys me ~22 extra minutes of deep sleep.

    The hard part has been doing meaningful pattern-finding with one person's data (n=1), on-device, without phoning home to a big model. Lots of tradeoffs there I'm still working through.

    Two things I'd genuinely love this crowd's take on:

    1. For an n=1 personal model how much history do you think people actually need before "it learns you" feels true rather than like a gimmick? I landed on 14 days but I go back and forth.
    2. Does "the AI runs entirely on your machine" read as a real trust feature to you, or as a limitation (no cross-device sync, no fancy cloud model)?

    Happy to get into the technical weeds if anyone's curious.

  3. 1

    Vitra's been a great pairing with my Oura Ring. The daily verdict sits right on both my work laptop and home desktop, so I've basically stopped reaching for my phone to check readiness. Didn't expect the deeper insights to be as useful as they are — caught a few trends I'd have completely missed otherwise. Really well built.

    1. 1

      Appreciate this. Removing that phone-glance friction so you actually sit with the data was the whole design goal, so it's good to hear it playing out that way. Thanks for giving it a proper run.

  4. 3

    I hope you like this. I built this for myself, but due to so much demand from fellow OURA users, I've decided to make it public and release it.

    Hope you enjoy it!

  5. 2

    The local-first angle is strong here because health data is one of the few categories where “no cloud” is not just a feature, it is part of the trust promise. Turning Oura data into a simple daily verdict also makes sense because most people do not need more charts. They need a clear read on whether to push, rest, or hold.

    I’d probably lean harder into that framing: private recovery intelligence, not another wellness dashboard. The product feels more like a calm decision layer on top of wearable data than a tracker.

    One thing to think about early is whether Vitra feels ownable enough if this expands beyond Oura into broader recovery, HRV, sleep, and readiness guidance. For a softer health/wellness direction, Lyriso.com would feel more natural and more brandable around daily guidance and recovery.

    1. 1

      Appreciate this! The "calm decision layer, not another dashboard" framing is exactly where my head's at. The whole reason it's local-first is that a recovery verdict only earns trust if the data never leaves the device, so I'm glad that read came through.

      On expanding past Oura — that's the plan (HRV, sleep, readiness across more wearables down the line), but the positioning stays the same: one clear daily read on push / rest / hold. The brand's built around that decision, not around any single device, so I think "Vitra" holds up fine as the surface widens.

      On that note: Strava and Withings integrations are already built and working. We're just holding the public release until they approve us past the single-seat dev limit.

      Thanks for taking the time to think it through.

Trending on Indie Hackers
Most founders don't have a product problem. They have a visibility problem User Avatar 98 comments Day 4: Why I Built a $199 Workspace Nobody Asked For User Avatar 52 comments How to automatically turn customer feedback into high-converting testimonials User Avatar 39 comments Spent months building LazyEats AI. Spent 1 day realizing I have no idea how to get users. User Avatar 32 comments Why Claude Skills Are Becoming Important for Tech Careers User Avatar 25 comments I kept rewriting the same quiz + spaced-repetition code. So I packaged it into an API User Avatar 21 comments