Things we have built, and why.

Three products for disc golfers, built on one shared foundation.

FairwayFlow is a community course wiki, and the map everything else reads from. Course information for this sport is scattered across half a dozen sites that all disagree with each other, so we built a pipeline that pulls from OpenStreetMap, the PDGA, Disc Golf Metrix and DGCourseReview, then merges, deduplicates and enriches. There are about 12,180 courses in the live database, 99.4 percent of them carrying a stable ID, which is the only reason records from four sources can be reconciled at all. Players verify each other’s contributions.

LogDog is the voice caddie. Say "Log Dog, birdie" and keep walking. It understands you in the cloud when there is signal and falls back to an offline parser when there is not, because a disc golf course is exactly where phones stop working. Skins, nassau, wolf and side bets ride along in the same flow.

Hyzerlyzer looks at your throw. Film a drive and it pulls a skeleton out of the video on your own device, scores the mechanics, then lays a pro’s skeleton over yours, scaled to your body so the comparison is about timing rather than height. It is in alpha and not in the app stores yet.

The suite as a whole is in alpha. What it shows is range: three genuinely different products on three domains, sharing a data layer, a ledger and a vision stack. That course count is measured. Everything else we would have to make up, so we will not.

If you want one example of what we mean by technology that brings people together, this is the one.

It is a matchmaking app where coupled friends build profiles for their single friends. No swiping. No algorithms. Just someone who actually knows you. Your friend writes your story, you approve every word before it goes anywhere, and the introductions come from people who already care how it turns out for you.

The idea came from a real moment. Chelsea Handler’s friend slipped a guy her number, they hit it off, and that is a thing good friends have always done for each other. We built the app that lets everybody’s friends do it, with a network of Handlers who compare notes and look out for their people.

We made it as a proposal alpha: a working thing rather than a pitch, built to show what this kind of software can be when the point is a real introduction instead of time spent staring at the app. It is live now, and still in development.

Arc is for students setting and monitoring their own goals. It started with a small unfairness.

Attendance systems know three things: on time, late, absent. A student who has been arriving twenty minutes late all term and gets themselves down to five minutes late has done something genuinely hard, and the system has no way to notice. It just says late again. Arc rewards the improving, not only the already arrived.

Students set bigger goals, break them into small ones and rate their own progress, while the classroom side quietly supplies the evidence. You check in by tapping your own photo, on time or late is worked out from the bell schedule, and hall passes run on a timer. The teacher gets a dashboard, somewhere to leave praise and redirect notes, and a follow-up list that surfaces the students nobody has spoken to in a few days, which is usually the quiet middle of the room.

It also had to be fun, or it would go the way of every paper goal sheet ever handed out: filled in twice, then lost in a binder.

The serious constraints turned into design features. Everything is anonymous, nothing links back to a named student, and who can see which goal is enforced down in the database rather than trusted to application code. It is hosted in Canada, with a self-hosted option for districts that need the data on their own network, plus an audit log, per-student export and delete, and a scheduled purge. The data was never at risk. We built to that bar anyway, because the users are kids and that is simply the bar.

Late in the build, a structured review of the code turned up five real security holes. Each one got a failing test before it got a fix.

There are no usage numbers to show you, and that is deliberate: there is no analytics in the product at all. It runs in two Career Life Education blocks at a Vancouver secondary school, with real rosters and attendance history carried over from the app it replaced.