AI engineering pilot · Hands-on consulting
Make one AI-assisted engineering task repeatable.
Work with me on a real task from your backlog. Your team leaves with a worked example, shared instructions and a clear way to check the result.
A paid, focused engagement. Scope, hours and availability agreed together.
The work
From an individual habit to a practice the team can use.
For teams already using AI coding tools that want help making one useful approach repeatable. We work with your engineers on a task they recognise.
- 01 / DEFINE
Choose the task and the checks.
Agree representative examples, expected behaviour and acceptance criteria with the owner. Review the available baseline, access and constraints before implementation.
- 02 / IMPLEMENT
Work through it together.
Try the task using your approved tools. Inspect generated changes, record corrections and review effort, then improve the repository instructions and checks.
- 03 / HAND OVER
Make it usable by someone else.
Have another engineer follow the example. Address agreed issues and hand the work to a named maintainer, with a recommendation to continue, revise or stop.
What stays with your team
Useful work you can keep using.
A worked example
A concrete example for the selected task, including a reviewed change or pull request where agreed.
Repository guidance
Instructions that capture the context, steps and review expectations for the next engineer.
Checks and recorded results
Acceptance checks, corrections and remaining limitations that support a decision about further use.
A clear handover
A named owner, maintenance notes and agreed next steps. Your team retains control of merge and release.
How we judge it
Can the team repeat the task and trust the result?
We agree what a correct result looks like before starting. Acceptance means the example can be repeated and passes the agreed checks.
We look at accepted changes, review effort and rework, using the evidence available. Usage and cost data can be included where relevant. The pilot informs a decision about this task; broader productivity gains need further evidence.
Who you work with
Directly with Matt Drankowski.
I lead AI-SDLC transformation in a Fortune 100 environment supporting 10,000+ engineers. My work combines developer enablement, shared AI practices, custom agents and delivery measurement.
I bring that experience to the agreed task and work alongside your internal engineering owner.
Read the enterprise adoption story →Working together
A clear scope before work begins.
Agree the practical details.
We confirm deliverables, acceptance criteria, rate, hour cap, billable meetings, access and availability in writing. Your team provides an engineering owner and time to try and review the work.
Keep the first step manageable.
I flag blockers and changes before using additional hours. If we reach the cap, work pauses with unfinished items made explicit. Additional teams or repositories are a separate scope. No prior assessment is required.
Start with your backlog
Which task would you like to make repeatable?
Tell me about the task, your team and preferred timing. We can use a 15–20 minute conversation to establish fit, then agree a paid scope.