A running app for people who still don't call themselves runners.
The hard part of getting into running is rarely the workout. It's fear, inconsistency, pain, and the sense that every other fitness app expects you to be further along than you are. OryxFitness meets people earlier than that.
Adaptive weekly plans respond to effort, soreness, and missed sessions instead of pretending life never interrupts
The product favors reassurance and clarity over leaderboards, streak panic, or pace-shaming
Voice coaching, GPS tracking, and bilingual support make the experience useful in the moment, not just pretty on a landing page
Currently in production testing — the app is not publicly live yet
Running apps are built for runners. The leaderboards, pace targets, streak mechanics, and social features all assume you already identify as someone who runs. For people who don't — who are scared of starting, inconsistent, dealing with pain or self-doubt — every existing app is a reminder that they're behind.
OryxFitness meets people before they call themselves runners. Adaptive weekly plans respond to effort, soreness, and missed sessions instead of pretending life never interrupts. The product favors reassurance and clarity over leaderboards and streak panic. Voice coaching, GPS tracking, and bilingual support make it useful in the moment — not just a planning tool but a present companion.
Adaptive plans that genuinely respond to human inconsistency required modeling recovery, soreness, missed sessions, and effort levels as first-class inputs, not afterthoughts
Voice coaching that sounds encouraging without sounding patronizing is a tone problem — the prompts were adversarially tested against early adopters who specifically identified as not runners
GPS tracking on mobile web needed to handle tunnels, poor signal, battery drain, and the fact that new runners often walk-run
A running product that people who don't call themselves runners are willing to open. The app is in production testing rather than public release — the adaptive planning model is being run against real inconsistency to confirm it bends without punishing users or breaking the progression.