Notes from the workshop.
Writing about building a drone mapping company the hard way — local AI infrastructure, autonomous daemons, operational rigor, and the specific technical choices we're making pre-revenue. Short posts, concrete details, no marketing voice.
The flight plan is the product
Our aircraft flew its first full mapping mission — 2,009 RTK-tagged frames over a Cumberland Plateau farm — and we processed the data three different ways. The mapping products came out survey-grade. The photorealistic 3D came out as a field of needles. The difference is the lesson that now shapes every quote we write: you can't fix capture geometry in post.
Read the post →The aircraft is the cheap part
This week we bought our survey aircraft — knowing it sits on a clock. We bought it anyway, on purpose, and none of the reasoning touches the thing you actually pay us for. A note on what's expensive and what's cheap in this business, and why your data is built to outlive any drone we fly.
Read the post →Building the operator before building the operation
Most drone shops at our stage are a laptop and a DJI and good intentions. We spent the pre-revenue runway building the layer behind the mission instead — a local AI stack, autonomous daemons, a mission control dashboard, and a nightly red-team framework that attacks our own system while we sleep. Here's why, and what it looks like.
Read the post →