A coach with context
Ask about your own training: volume, recent sessions, laps, heart rate, and the week ahead. Answers draw on synced or uploaded runs instead of starting from a blank conversation.
← Back to selected work
Product leadership & engineering
Trackside Training.
An AI running coach, built around the training your watch already records.
From product direction to the code, integrations, and releases: my work as co-founder and technical lead of Fair Weather Athletics.
- My role
- Co-founder & technical lead
- Product
- AI running coach
- Platforms
- iOS, Android & web
- Company
- Fair Weather Athletics
The problem
A plan has to fit
the runner.
Training is a moving target. A race gives you a deadline, but your available days, recent runs, and day-to-day constraints determine what a useful plan should look like.
Trackside brings that context into one place: the training an athlete has already recorded, a coach they can ask about it, and a plan they can review and adjust.
My responsibility: the whole product.
I built the platform across Flutter mobile, React web, FastAPI services, and AWS, along with the device integrations, operational tooling, and release paths that connect them. The work spans product decisions, interface design, architecture, implementation, and ongoing delivery.
The athlete experience
Connected training.
A useful conversation. A practical plan.
A plan you can review
Build toward a race with a plan that considers your available days and training. The first plan arrives as a draft to review and accept. AI-built plans are a Premium feature.
Room for life to change
Move a session, shorten a workout, or discuss a change to the week through the coach. Plan changes by chat are Premium; proposed week resizing is applied once the runner agrees.
Designed around existing habits
Keep the watch.
Connect the training.
Trackside connects with Garmin, Suunto, Wahoo, Apple Health, and Health Connect, with file upload available for recorded activities. Device sync begins with the next sync after connection; Apple Health can bring in up to 180 days on the first sync.
Training-aware coaching is free and needs synced or uploaded runs. AI-built plans and plan changes by chat are Premium features. The product keeps those distinctions visible in the experience and its marketing.
Explore the product →What I built
One product.
Several connected systems.
The coaching interface is the visible part. A shared data model, dependable integrations, background work, and delivery systems make it usable across devices.
Flutter · iOS & Android
Mobile experience
Coaching, calendar, training insights, health-data connections, and the mobile release process.
React · TypeScript
Web experience
A browser home for athletes, with organization and coach workflows alongside the training experience.
Python · FastAPI · MySQL
Coaching & data
Training-data ingestion, background processing, coaching tools, and plan generation behind a shared API.
AWS · Integrations · Releases
Operations & delivery
The cloud infrastructure, deployment paths, device connections, and operational tools behind the product.
Why this work matters
Engineering ownership from the first decision to the next release.
Trackside is an example of how I work across a product: understand the user, make the tradeoffs, build the systems, and stay responsible for how they fit together.
Get in Touch
Building a product with this many moving parts?
Tell me what needs to work together, what you have already built, and where you need experienced technical ownership.