Services
Four ways to work together
Most engagements start with a diagnostic audit, then convert into development work once we both know what needs doing. A training course brings your team up the same ladder, and an ongoing advisory retainer keeps the relationship going between larger pieces of work. The course and the retainer have a fixed shape and a stated price, both on their cards below. The audit and the development engagement are scoped to your situation, so their number comes out of a conversation about specifics, not a rate card.
Audit / assessment
A fixed-scope diagnostic engagement. I go deep on your codebase and its build and test setup, then hand back a written report: what the real problems are, and a roadmap of what to do about them, in priority order.
It stands on its own as dedicated expert work, and it is the low-commitment way in. When it makes sense, the roadmap becomes the plan for a development engagement, so you see how I work before committing to anything larger.
Training course
A structured three-day course that teaches the whole verification ladder to your team. Day one covers the agent-ready foundation: toolchain, CI, and the context an agent needs. Day two is testing without hardware, meaning compile-time tests and mocking. Day three builds the software-in-the-loop harness and the full agent loop, ending in a capstone your engineers write themselves.
It runs on training code rather than your production codebase, so your team can experiment and get things wrong with nothing customer-facing at risk, and there is no production repository to expose or set up for a three-day course. When it helps, it extends with extra days that re-run the ladder against your own microcontroller family, answering the "does this work on our hardware" question directly.
Development
A milestone-based development contract, typically running three to six months. Each milestone has acceptance criteria agreed up front (for example, refactoring a specific module and landing a test suite around it), so progress is measured by what actually ships, not by hours logged.
This is where the verification loop gets installed directly in your own codebase, not demonstrated on training code.
Ongoing advisory
Roughly ten hours a month of ongoing access. Design reviews, second opinions on architectural calls, help when your team hits something outside its usual ground.
It suits teams who have delivery under control and want a steady line to someone who has seen the failure modes before, without standing up a full engagement each time.
Not sure which fits?
If you can describe the problem but not the shape of the fix, start with the audit; that's exactly what it's for. If you already know what needs building, say so and we can scope a development engagement directly. The course and the retainer already carry their price, so there's nothing to quote there: pick one and get in touch. Either way, the first step is the same: tell me what you're dealing with.