SHORT. SPECIFIC. HONEST.

Most of the work in the last few months has been deletion. Pre-Trailhead AI scaffolding came out. Streaming RPCs we never finished came out. Whole services that looked good in a PRD but never earned their place in the product came out. Tens of thousands of lines, gone.

What's left is the loop we kept coming back to: groups, activities, a feed scoped to people you actually know, a map you can rely on, and Trailhead running on the device. That's the founding-beta build.

Some of what came out lived in earlier iterations of this product under different names. Some of it lived in the original PRD. None of it earned its keep against the version of Nomatic Atlas we're shipping. The road ahead is shorter, on purpose.

We will keep cutting. If a feature can't explain why it's there, it doesn't get to stay.

We started chasing this idea nine years ago. It looked different then — different name, different scope, different stack. The thing that hasn't changed is the problem we were trying to solve.

Every outdoor app we tried did one of two things. It turned a quiet morning on a river into a content opportunity, or it turned the people we trusted into a follower count. Neither felt right. Both felt loud.

What we wanted was simple. A place to plan a Saturday hike with three friends without spinning up another group chat. Groups that didn't have to perform for an audience. Maps that worked outside. An assistant we could ask about a fork in the trail without sending the question to a server farm.

We've rebuilt this product more times than is reasonable. The version that ships in the founding beta is what's left after nearly a decade of cuts. We took the long way on purpose.