Applied research - taking observability into domains it hasn't reached yet.
Our founder is running a public fitness transformation instrumented like a production service. Daily body and nutrition metrics scraped into Prometheus. A training diary in Loki - every workout, meal deviation and missed session a structured log event. Workouts traced with OpenTelemetry: each session a trace, each exercise a span, each set a span event - and the rest periods are the gaps in the waterfall.
Fitness goals are SLOs with real error budgets: multi-window burn-rate alerts fire when missed sessions burn the budget too fast, and plateaus get blameless postmortems with root causes and action items. If we'll instrument ourselves this thoroughly, imagine what we'll do for your platform.
An OpenTelemetry reference architecture for smart home gyms and connected fitness equipment. A collector at the edge ingests per-rep data from equipment APIs and BLE sensors - load, tempo, range of motion - and auto-generates workout traces using semantic conventions for reps, sets and sessions. No manual logging, no proprietary silo.
The value is in correlation: training telemetry joined with nutrition and recovery signals - calories, protein, sleep - to surface performance insights no single device can see. Did output drop because of the programme, or because of three short nights and a low-protein week? It's the same correlated-signals question we answer for production systems, asked of the athlete.
For equipment manufacturers and fitness platforms: open, standards-based telemetry makes your hardware a first-class citizen in every customer's data stack - richer insights for your users, integrations without bespoke engineering, and a genuine differentiator on the spec sheet. We're looking for partners to build the reference implementation with. Talk to us.