A hardware product that needed a brain.
Equitek came to IVI with a hardware-first problem: they were building a wearable sensor patch for racehorses to monitor health and performance data in real time. The sensor was the product — but they needed a companion app to make that data meaningful to the people managing the horses.
They had a target market — racing and stable operations teams in the UAE, particularly Dubai — and wanted a prototype to demonstrate the concept to investors before the hardware was production-ready. There was no existing design, no validated feature set, and no design precedent to borrow from.
The real challenge: Design a credible, detailed health monitoring experience for a domain where I had zero prior knowledge — well enough that domain experts would trust it.
Learning the domain before touching Figma.
The first thing I did was a structured discovery session with the Equitek founders — not to gather user research, but to understand what they already knew: target users, competitive landscape, technical constraints of the sensor, and which metrics they considered essential.
I then mapped the existing equine monitoring market — Equimetrics, HorsePal, Hoofstep, Steed, Horsano, and Arioneo. Most were built for vets and data analysts, not stable staff making fast decisions under pressure. That gap — accessible, actionable monitoring for operations teams — became a core design principle.
Key insight: The goal of the app was not to show data — it was to show whether the data was normal for this moment. That distinction drove every information hierarchy decision.
Building a reference system from scratch.
Using AI-assisted research and secondary sources, I built a detailed understanding of equine health metrics — structuring data into reference tables to make real design decisions about what to display, what thresholds trigger alerts, and how to visualise performance across a race session.
| Vital Metric | Normal Range | Alert Threshold | Why It Matters |
|---|---|---|---|
| Heart Rate | 40–180 bpm | Alert above race zone | Primary indicator of exertion and cardiac stress |
| Body Temperature | 37–38.3°C | Alert on recovery slowdown | Flags heat stress and infection risk |
| Blood Glucose | 85–115 mg/dL | Alert on fatigue-related dips | Energy availability and metabolic stress |
| Oxygen Saturation | >94% | Custom recovery thresholds | Respiratory efficiency and endurance capacity |
| Gait / Movement | Symmetry index >90% | >10% asymmetry = alert | Early detection of lameness or injury risk |
Gait analysis added another layer of complexity. I mapped gait types to race phases so the app could surface contextual warnings — not just raw numbers.
| Gait Type | Description | Speed Range | Race Phase |
|---|---|---|---|
| Walk | 4-beat, slow and steady | 4–6 km/h | Pre / post race |
| Trot | 2-beat, diagonal pace | 8–12 km/h | Warm-up (Q1) |
| Canter | 3-beat, smooth | 16–27 km/h | Build-up (Q2) · Cooldown (Q4) |
| Gallop | 4-beat, fastest gait | 40–65+ km/h | Peak intensity (Q3) |
Three modes of use. One coherent system.
With enough domain knowledge to make intelligent decisions, I moved to IA. The core question: what does a racing operations team need to see, and in what order?
Live monitoring — real-time vitals during a race or training session, with threshold-based alerts surfaced immediately.
Session review — post-session analysis across Q1–Q4, showing how each metric tracked against expected ranges for that phase.
Horse health history — longitudinal records per horse, enabling trainers to spot patterns and make informed decisions about training load.
IA was validated with the Equitek founders before wireframing. Their feedback removed out-of-scope features and sharpened the focus on the core monitoring experience.
Six screens. A complete product vision.
The prototype covered the full core experience — from a high-level fleet view to individual horse session breakdowns and alert management. Every screen answers a specific question a racing operations team member would have in a specific moment.
The silence said everything.
When I presented the prototype, the founders went quiet for a moment before responding. They had not anticipated the level of detail — they had expected something much more surface-level.
"They had not thought about the app at such a detailed level and were not expecting this."
— Equitek founders, on seeing the prototypeThey requested removal of a small number of screens outside their immediate scope. The core monitoring experience was approved without significant changes.
From zero domain knowledge to investor-ready.
A full high-fidelity prototype built from zero domain knowledge through structured research, competitive analysis, IA validation, and iterative design. The prototype serves as Equitek's demonstrable product vision for investor conversations — and also functions as a specification document for what data the sensor needs to capture.
The biggest gap was the absence of actual user research with racing operations teams. The domain research was thorough — but it was research about horses, not research with the people who manage them.
The founders were the proxy for the user throughout, which worked for a first prototype but would not be sufficient for a product moving toward real deployment. If there is a Phase 2, the first thing I would push for is access to a few operations team members in Dubai for even brief interviews.