All work

Concept · Equine data · 2026

Thoroughbase

A mobile-first virtual barn that turns a racehorse’s RFID chip into one complete, permission-aware record.

Type
Self-initiated concept
My role
UX Researcher · Product Designer
Platform
Mobile & tablet, mobile-first
Tools
Figma Make (AI prototyping)

01

Overview

A thoroughbred racehorse generates a remarkable amount of data: vet visits, X-rays, race results, bloodlines, treatment plans and, more and more, live readings from RFID chips and wearable sensors. Today that information is scattered across paper files, vet portals, track systems and third-party apps, and the people who care for a horse each need a different slice of it.

Thoroughbase is a concept for a mobile-first “virtual barn.” It links everything known about a horse to the RFID chip already implanted in its neck, pulls in wearable data automatically, and shows owners, trainers, managers and foremen only what’s relevant to their role.

About this project. Thoroughbase is a self-initiated concept. I wrote the product brief and generated the interactive prototype with Figma Make. The research below covers the desk research, stakeholder mapping and heuristic review I’ve done, plus the plan I’d use to validate the concept with real barn teams. Personas are assumption-based proto-personas, clearly marked as hypotheses to test.

02

The challenge

How might we give everyone in a racing barn one trustworthy source of truth for each horse, without overwhelming them or exposing data they shouldn’t see?

Three tensions shaped the problem from day one:

  • Completeness vs. focus. The goal is to hold all of a horse’s information, but a foreman at 5 a.m. needs one or two things, fast.
  • Connected vs. private. Integrating chips and wearables makes data effortless to capture, and makes permissions essential. Medical and financial records are sensitive.
  • Serious vs. delightful. The brief asked for something that feels “almost like a game, where the horse is the sim,” while still handling medical history with the gravity it deserves.

03

Objectives

Unify

One living record per horse: medical history, X-rays, performance and accolades, pedigree and breeding, treatment and maintenance plans, and notes.

Connect

Use the horse’s RFID chip as its key, and bring wearable data into one dashboard so readings flow in instead of being re-keyed.

Protect

Role-based profiles and assignable permissions for owners and partners, trainers, managers and foremen, so sensitive data stays with the right people.

Delight

Make daily check-ins glanceable and rewarding, “caring for your sim,” so the app is used every day, not just at vet visits. Cool and premium, not cartoonish.

04

Research approach

Because this began as a concept, I front-loaded the research I could do without a client, and designed the studies I’d run next with real barn teams.

  1. Desk research. How RFID and wearables are used in racing today, and what data already exists for each horse.
  2. Stakeholder & role mapping. Who touches a horse’s data, what each role needs, and what they should never see.
  3. Proto-personas & jobs-to-be-done. Assumption-based personas for the four core roles, to make the hypotheses explicit and testable.
  4. Prototype & heuristic review. Generate a working prototype with Figma Make, then evaluate it against usability heuristics.
  5. Validation plan. Field studies and usability tests with trainers, foremen, managers and owners to confirm or kill each assumption.

05

What desk research told me

RFID in horse racing typically works on two layers:

  • Passive microchips implanted in the horse’s neck carry a unique ID.
  • Active sensors on equipment, such as saddle cloths, track identification, speed, position and safety metrics.

Readers placed around a track enable precise automated timing, fraud prevention and real-time monitoring of 40+ horses at once. Meanwhile, a growing category of equine wearables works like a fitness tracker for horses, capturing heart rate, temperature, activity and recovery.

The opportunity wasn’t a lack of data. It was the lack of one place, organized around the horse, that ties these streams together. That became the core of the concept:

06

Who uses it

Mapping the roles in a barn turned a vague “permissions” requirement into a concrete access model. This matrix became the backbone of the Team screen.

DataOwner / partnerTrainerManagerForeman
Horse profile & statusFullFullView allAssigned horses
Medical records & X-raysFullEditViewStatus only
Treatment plansFullManageViewUpdate daily progress
Performance & race historyFullFullReports—
SchedulesViewEditManageView assigned
Financial recordsFull———
Team & permissionsAdmin———

Proto-personas

These are hypotheses, not research findings. Each one names an assumption I’d validate in the field.

Trainer

Sarah, the trainer

Juggles training plans, vet guidance and race prep for every horse in her care.

Needs: health and performance trends at a glance; fast access to the latest treatment note.

Assumption to test: trainers will check the app between sets during morning works.

Owner / partner

Michael, the owner

Invested in several horses and wants confidence that they’re healthy and progressing.

Needs: race results, earnings, costs and a clear health summary.

Assumption to test: owners care most about performance and spend, and less about daily logs.

Manager

Jessica, the barn manager

Keeps schedules, staff and records running across the whole barn.

Needs: a view of every horse, scheduling and exportable reports.

Assumption to test: reporting is a weekly task, so it can live a level deeper.

Foreman

Marcus, the foreman

Hands-on with assigned horses all day, often with gloves on and little time.

Needs: scan a horse, see today’s tasks, log status in seconds.

Assumption to test: foremen will scan before they search, and need big, simple controls.

Jobs to be done

  • When a horse comes in from morning works, I want to log how it went in seconds, so the trainer sees it before deciding on afternoon care.
  • When the vet updates a treatment plan, I want everyone assigned to that horse to see the change, so no one follows yesterday’s instructions.
  • When I scan an unfamiliar horse, I want its identity and current status confirmed instantly, so I know I have the right animal.
  • When I’m reviewing my investment, I want performance and health in one summary, so I can make decisions without chasing updates.

07

Design principles

The horse is the hero

Every screen starts from a horse, like a character in a sim: a profile, live vitals, history and a family tree.

Right data, right person

Permissions shape the interface itself, so each role sees a focused version of the same truth.

Glanceable first

Scores, status chips and live indicators up top; detail one tap away. Built for one hand, outdoors.

Premium & calm

A dark, gold-accented palette that feels like race day, keeps glare down in early mornings, and treats medical data seriously.

08

The prototype

I turned the brief into a high-fidelity, mobile-first prototype with Figma Make, then explored it screen by screen. Partner device names in the prototype, like EquiTrack Pro and HorseVital Plus, are placeholders.

Barn home screen: welcome message, stable summary, counts of active, training and recovery horses, and horse cards with live indicators
Barn. The virtual barn: stable status at a glance, a live indicator for horses wearing sensors, and a Scan button for the RFID reader.
Horse profile for Thunder Strike with chip number, health score 95 percent and performance 88 percent, and basic information
Horse profile. The chip ID sits under the name. Health and performance scores lead, with live wearable data below.
Health dashboard with average health, connected devices, a recovery alert and wearable readings for heart rate, temperature and stress
Health. Alerts surface horses that need attention, next to live readings pulled from wearables.
Team screen listing staff with roles and permission chips
Team. Each person has a role and visible permissions, straight from the access model.
Medical tab with a checkup and vaccination record, diagnosis, treatment and follow-up date
Medical. Checkups and vaccinations with diagnosis, treatment and follow-ups.
Performance tab showing race results with finishing position, track, distance, time and earnings
Performance. Race history with placing, track, distance, time and earnings.
Treatment tab showing a pre-race conditioning plan with progress and the assigned foreman
Treatment. Plans with progress and an owner, so daily care is accountable.
Family tab showing the sire, dam and heritage of the horse
Family. Pedigree and breeding history, a big part of a thoroughbred’s story.
Records screen with medical, performance and X-ray filters and an upload button
Records. One searchable history across medical, performance and X-rays.
Settings screen listing connected devices and preferences
Settings. Where third-party wearables and the RFID scanner connect.

09

Heuristic review: what I’d fix next

AI gets you to a testable artifact fast, but it doesn’t do the judgment for you. I reviewed the generated prototype against Nielsen’s usability heuristics. These issues are real, and they set the agenda for the next iteration.

FindingHeuristicNext iteration
Scan, the hero action, opens a placeholder alert instead of a scanning flowVisibility of system statusDesign the full flow: scanning state, haptic confirmation, “horse found” card, manual search fallback and an unregistered-chip error
Device status labels (Connected / Not Connected / Configured) are clipped off the edge of the Settings screenVisibility of system statusMove status under each device name as a labeled chip that wraps safely
Vet names render as “Dr. Dr. James Peterson”Consistency & error preventionStore titles separately from names and format them in one place
Health and performance appear as bare percentagesMatch between system & real worldExplain what drives each score, show the trend, and link to the underlying readings
Permissions are visible on the Team screen but not editableUser control & freedomAdd an edit-permissions sheet with presets per role and a clear review step
Small gold text on dark surfacesAccessibilityAudit every text and icon pair against WCAG AA and adjust tokens where needed

10

Validation plan

The next step is getting this in front of the people who’d use it. Each study maps back to an assumption above.

Contextual inquiry

Observe morning works and evening rounds at two or three barns to see how information is captured and shared today, and where it gets lost.

Card sort

Have trainers, managers and vets sort record types to validate the Records taxonomy and the horse-profile tabs.

Task-based usability tests

Scan a horse and confirm its identity; log a treatment update; invite a foreman with limited access; find the latest X-ray.

Diary study

A week with foremen logging daily status, to test whether quick, game-like check-ins hold up in real barn conditions.

Success measures I’d track

These are targets to validate, not results:

  • Time from scan to the right horse’s profile: about 3 seconds
  • Foreman logs daily status in under 30 seconds
  • Trainers find the latest treatment note in one tap from the horse profile
  • Zero permission leaks across role-based test scenarios
  • Daily active use by barn staff, not just at vet visits

11

Reflection

Thoroughbase let me practice the part of UX I love most: turning an ambitious brief into clear, testable bets. AI prototyping got me from idea to an interactive app in a fraction of the usual time. That speed is only valuable if research decides what’s worth building, so the real work starts with listening to the people in the barn.