Weekly Status Update: July 11th, 2026

Been a wild month with no updates. I started by attending WWDC in June. It’s great to be in the Apple community, chat with folks, learn about what they’re up to, and meet friends. I had the opportunity to present Dewy Golf at Indie Dev Fair (CommunityKit conference), which was a blast. I showed off v2 experiences, and how the vision of current solutions will expand. I’m resolved (I hope) to restart these updates, as they’re a great way for me to share and look back.

Let’s get to updates.

Maps

I’ve built out solutions for most/all countries I plan on supporting. Obtaining mapping data is a patched-together network of providers and cost. I’ve got the pipeline doing detections easily. I can process about 700-2000 courses an hour, depending on the map data source.

Maps continue to be the biggest lift of this project. What I expected was going to be 3 weeks has turned into a ground fight lasting over 4 months. My current task is continued work building out my conflation and spatial reasoning pipeline. It’s coming along, feel like it’s close, but we shall see.

Dewy Golf (iOS/Watch)

I spent a lot of time locking in the bugs that were existing in the watch app/sync. I think I’ve verified with tests that we are in a much better place. I’m now working on final UI. I have a new method to create your own custom UI that meets the individual golfer’s needs. Sometimes these golf apps (workout apps in general) are packed with information, which makes them less glanceable.

Dewy Golf (Server)

I’ve recently done an audit of my server security, privacy protections, and anti-scraping. I was already pretty good, but had left open some doors for myself/backend systems that for launch really should be closed. They are closed now.

ML Training

Continued collection, validation, and working on both the light and mobile models at the same time. Given Apple Watch size, battery, etc., it’s better to have two models (technically 3) for the watch, the phone, and client (my) side. I’ve added in the flow to allow users to record and I post-process it on my side. Hoping that gets folks more interested in recording, gives me data, and they get the full closed loop.