A track-day notebook for motorcycle riders that knows the circuit: 38 real tracks with a computed reference line, a session logbook built for one-handed use between sessions, and a Mac companion for the analysis that happens the day after. I owned the whole thing, from identifying the rider and the problem, through 16 interviews and a TestFlight beta, to the SwiftUI build and the App Store launch — in one month, as a solo founder working with AI.
Apex Lines is the study-and-logging half of a motorcycle track day. Before the day, a rider walks the circuit on a real map with numbered corners, a computed reference line, and per-corner guides. Between sessions, they log lap times, tire pressures, weather, and corner notes with one hand. The day after, the same notebook opens on the Mac as a three-pane logbook with lap trends, personal bests, and a Garage that turns logged ride time into engine hours and tire heat cycles. Offline-first, no account, one subscription across iPhone and Mac.
I ride track days myself, and the app I wanted did not exist. The incumbents are live lap timers built around car HPDE: they want your attention while you are on the bike, they assume a mounted phone and hardware, and their notes are an afterthought. A rider in a paddock has gloves on, a session in twenty minutes, and a logbook that is still paper. The brief was to design for the moments a lap timer never sees: the night before, the ninety seconds between sessions, and the day after.
“Riders want to spend their time riding. The app's job is to be brilliant before the day and after it, and almost invisible during it.”
The idea started as my own frustration, so the first job was to find out whether it was only mine. I began with casual paddock conversations to test one assumption: that there was no high-quality logbook built for motorcycle riders. Once that held, I ran 16 structured interviews with riders across the experience curve, from first-track-day novices to advanced group regulars, 9 conversations with track-day coaches, a rider survey for qualitative feedback at scale, and several weeks of reading how riders actually talk about tools in track-day forums and groups.
Three findings shaped everything that followed. The market is car-first. Almost every app riders had tried was designed around car track experiences, and the motorcycle-specific details (tire pressures per wheel, gearing, heat cycles, gloves) were missing or awkward. Riders want to ride, not tap. The meaningful time with an app happens before the track day and the day after it, not during. Trackside use is quick checks only. Between sessions a rider has a couple of minutes, one free hand, and no patience for forms.
Design for three moments, not one app. The night before gets planning, checklists, and study material. The paddock gets a one-handed logger that captures a session in under ninety seconds. The day after gets a full logbook on the Mac with charts and setup recall. Same data, same components, a different shell for each moment. This one decision set the information architecture, the platform split, and the free tier.
The first user interview was with myself. Every later one was with a rider who had the same paper logbook in the van.
A verified competitor teardown of RaceChrono, Harry's LapTimer, and APEX Pro confirmed the wedge: every incumbent is a live lap timer, and nobody owned pre-ride study or the logbook. Apex Lines would never claim to be a faster timer. It would be the tool your lap timer is missing, motorcycle-first everywhere, and honest about being study material rather than coaching.
The business model followed the research rather than the other way around. Freemium subscription apps convert a small single-digit share of installs, which vetoed paid acquisition and made a genuinely useful free tier the growth engine: every circuit and every reference line free, the first track day free, Pro for the 3D fly-through, corner guides, and the unlimited notebook. I also committed to a privacy promise the audience cares about: no account, no analytics SDK, attribution through App Store campaign links only. That promise shaped the launch plan as much as the product.
What stays free: every circuit, every numbered corner, and the computed reference line on the 2D map. Pro opens 3D, fly-the-line, and the corner guides.
I built Apex Lines alone, with Claude as a collaborator in every phase. This is worth being precise about, because it is the reason a designer could take an idea from research to the App Store in a month without a team.
Research and market analysis. I used Claude to synthesize interview notes and survey responses into themes, to run a multi-source market investigation, and to adversarially verify every number in the launch plan against primary sources. Claims that did not survive verification were marked as unverified and never made it into a decision. Design system. The brand palette, type scale, spacing, and shared components (readouts, steppers, condition chips, the timecode field) were generated as SwiftUI code from my design direction, reviewed, and refined in Xcode previews rather than in static mockups. Prototypes and build. Each milestone was specified by me as a written design spec, built with Claude as pair programmer, and verified with simulator screenshots and deep-link QA routes before the next one started. When two proposals conflicted, I resolved them with a documented pick, not a hedge.
The division of labor stayed the same throughout: I owned the rider, the problem, the priorities, and every design call. AI collapsed the distance between a decision and a working screen from days to minutes, which meant more of the month went to testing with riders and less to production.
The paddock persona set the interaction rules for iPhone: thumb reach, large targets, haptic confirmation, and undo everywhere. Lap times are entered in a stopwatch-style timecode field where digits fill in place, so a rider types 14732 and reads 1:47.32 without ever seeing a keyboard modal. Tire pressures and gearing use steppers instead of text fields. Corner notes are pinned by tapping the corner on a mini map, not by typing its name. Every motion token is gated on Reduce Motion.
The visual language borrows from the pit wall: racing red for the reference line, apex amber for the point that matters, a monospaced readout face for numbers, and dark ink surfaces that survive direct sunlight. The app icon, a late-apex line sweeping a single corner, is generated from the same geometry the app draws.
Track-day prep: steppers for rotation and pressures, and a countdown that reads at arm's length
The Garage: monospaced readouts on dark ink, amber only for the number that matters
38 real circuits, every corner numbered
Fly the reference line in 3D with a live HUD
Every corner briefed: brake, turn-in, apex, exit
38 real US circuits from OpenStreetMap geometry, with numbered corners and a reference line solved from the pavement itself by minimum-curvature optimization. Apexes are geometrically real, not hand-drawn, and the line can never leave the asphalt.
Track days hold sessions, sessions hold laps. Cold and hot pressures per wheel, weather auto-filled from Apple Weather, per-corner notes, lap trend charts with personal bests, and setup recall the next time you return to a circuit.
Every circuit renders as a 3D scene. Fly mode runs a physics-flavored lap over the line, braking into corners and driving out, with a live HUD. Each corner gets a guide with radius, run-up, and brake, turn-in, apex, and exit references derived from the line's curvature.
Prep checklists, run-group timers, overnight plans with a security checklist, and a shareable setup card. The Garage is a maintenance ledger keyed to logged track time: engine hours and tire heat cycles accrue from the sessions you already logged, with cited service intervals you accept or edit.
The logger is the product's answer to the second research finding. It opens as a bottom sheet with the lap field already focused, a number pad instead of a keyboard, and the previous session's pressures carried forward so the rider only edits what changed. Each saved lap gets a haptic tap; a new personal best gets a pulse.
Everything is undoable from a toast, so a gloved mis-tap costs nothing. When the rider comes back to the same circuit months later, the day header reads their last setup back to them: "Last time here: front 30.0 → 33.5." The notebook pays the rider back for filling it in.
The data model, units, charts, and most components are shared between iPhone and Mac. Only the shell and the input affordances diverge. On iPhone the notebook is a navigation stack tuned for one hand. On the Mac it becomes a three-pane logbook: a season sidebar, a lap table with deltas to best and previous, and an inspector for analysis and export, driven by keyboard shortcuts. iCloud sync carries a day logged trackside to the Mac by the time the rider gets home. The free tier never hides synced data: the cap is on creating new days, never on seeing your own.
Mac: season sidebar, day sheet with lap table, analysis and export inspector
iPhone: the same day, read one-handed
The first complete build went to a closed TestFlight beta of 9 riders drawn from the interview pool, so the people who had described the problem were the first to judge the answer. I asked each of them to use it for a real or simulated track day and to talk through the notebook the day after.
The beta did three things. It surfaced the initial bugs and rough edges that only real devices and real data expose. It drove a round of UI quality work: a custom icon set, transitions and animations on list inserts and chart reveals, and polish on the library cards, which went through four iterations before the map became the card. And it added a feature. Talking through the notebook with testers kept landing on the same paddock question, "how many hours are on this oil?", so the Garage, a maintenance ledger fed by the sessions riders were already logging, was designed and built during the beta and shipped in 1.0.
Beta polish pass: custom icons, motion, and the prep hub on Mac
Added during the beta: the Garage, a maintenance ledger fed by logged sessions
Launch was treated as part of the design, not an afterthought. Because the app collects no analytics, measurement had to be designed in: App Store campaign links per channel, offer-code batches per partner, and cookie-free redirect counting on the marketing site. The go-to-market leaned on zero-cost levers the research had validated, App Store optimization into the "racing line" and "track day notebook" gap, partnerships with track-day organizations already listed in the app, and a YouTube library of corner-by-corner track guides, with a single capped Apple Search Ads test as the only paid spend. The marketing site shipped alongside the app with written guides for 19 circuits and an in-browser racing-line lab as free, shareable entry points.
“First app that actually gets what a track day is for a rider. I punch in my laps between sessions with my gloves half on, and the next weekend it tells me what pressures I ran last time. My paper logbook is retired.”
“I send students to the corner guides before they show up. Being able to walk the line on a map and then see it fly in 3D makes the debrief so much faster. It is study material done right.”
“The month was not fast because I skipped steps. It was fast because every step produced a working thing riders could react to, and AI made the gap between a decision and a screen almost disappear. The judgment still had to be mine.”