Skip to content
BRIEFSHIPPEDMODEL LAYERAPPLICATIONDATA & EVALS


We're a software and AI studio. Most projects fail in the decisions made before anyone writes code, so that's where we start.

What we do

Two tracks. Often one project.

Software

Product engineering

Web and mobile products built to be handed over, not babysat. Typed end to end, tested where it counts, deployed on infrastructure you can afford at your current size.

Detail

AI

Applied AI systems

Language models wired into real workflows, with the evaluation harness built first so you can tell whether it's getting better or just getting different.

Detail

Both

Technical advisory

A second opinion before you commit. Architecture reviews, build-versus-buy calls, and honest answers about whether AI is the right tool for the problem you described.

Detail

How it goes

Four steps, in order, every time.

01

Scope

A week, fixed fee. We map the problem, the constraints, and what success actually looks like. You leave with a written plan whether or not you hire us.

02

Draw

Architecture, data model, interfaces. Decisions get made on paper where they're cheap to change, and written down where you can argue with them.

03

Build

Two-week cycles, deployed at the end of each one. You see working software throughout, not a demo at the finish line.

04

Hand over

Documentation, runbooks, and a session with your team. The goal is that you don't need us. If you want us anyway, that's a different contract.

Tell us what's broken, or what doesn't exist yet.

First reply within one working day. The scoping week is fixed fee, and you keep the plan either way.

Start a project