Seattle, WA
About
From dirt to tokens.
We research the entire AI compute stack.
Every layer between raw land and running inference is broken or bottlenecked — siting and permitting, data center buildout, grid interconnection, and the clouds that turn power into intelligence. Last Myle Engineering researches that entire lifecycle: what's actually happening in each market, where the constraints are moving, and what it takes to win at each layer.
We are a team of humans and agents
A small human team directs a large staff of AI agents. Our agents run the research pipeline — monitoring signals from land markets to model serving, running simulations, and synthesizing findings across the stack — while humans make the calls that matter. We operate our own company this way, and we build this capability for our clients.
What we do
We don't buy land, build data centers, or operate clouds. We do the work that makes all of those easier to understand:
- Market intelligence across the stack. Continuous signal monitoring from land markets and permitting politics to grid rules and compute pricing. We know where the moratoria are, when they expire, and which model ordinances are becoming the template.
- Simulation and scenario modeling. We model each layer of the lifecycle before capital is committed: siting outcomes, build timelines, interconnection queues, compute economics. Decisions get stress-tested in silico, not on the balance sheet.
- Experiment design and execution. Rapid, rigorous experiments on the questions that decide projects: market entry, timing, technology choices. We design them, run them, and hand you decision-grade results.
- Research staff on demand. A deep bench of university students we train and direct, scaling analytical capacity up and down with your project.
How we work
- Week 1: Our agents map your position on the dirt-to-tokens spectrum and identify the binding constraint. No cost.
- Week 2–3: Humans and agents run the experiments and simulations — in your context, your jurisdiction, your risk profile.
- Week 4: You get a decision-grade playbook and the option to have our research staff operate alongside your team.