Native engineering.
Better AI workflows.
I help teams build reliable iOS and macOS products and make AI useful in their everyday work. Start with a real task or engineering problem and agree what a useful result looks like.
AI workflows for your team
Help your team get more value from AI in its everyday work. I work alongside you to identify useful applications, improve existing workflows, and build shared practices around context, tools, and checking results. We start with real tasks and leave your team with workflows it can understand, evaluate, and maintain.
What we work on
- Identify where AI can help and where it adds overhead.
- Build reusable instructions, skills, and automations around real work.
- Design verification loops that check results after builds and feed findings into the next iteration.
- Help people develop confidence through hands-on sessions.
What you take away
A working example, reusable guidance and tools, verification checkpoints, and a team walkthrough. Evaluate time spent, review effort, and quality before deciding what to adopt more widely.
Focused engineering review
A persistent reliability issue, a delivery bottleneck, or an architecture decision that needs another perspective. I work with your team to understand the system, examine the evidence, and identify the changes that matter most.
What we look at
- The problem, constraints, and existing diagnostics.
- The relevant code and the way the system behaves.
- The main failure or bottleneck, and options for addressing it.
What you take away
A concise explanation, evidence behind the findings, prioritized recommendations, and a walkthrough with your team. A small reference change can be included where it fits the agreed scope.
Implementation & stabilization
When the team needs experienced hands to make a change or deliver a difficult feature. I can help with on-device AI, concurrency, audio and transcription pipelines, module boundaries, and build or release workflows. At Littlebird, I built the on-device meeting transcription pipeline for its macOS meeting assistant.
When it fits
You have a bounded engineering problem, a difficult native feature, or an agreed direction that needs implementation and technical leadership.
What you take away
The agreed change, relevant verification, integration with your team’s workflow, and a clear handover of the decisions and remaining limitations.
How we start
-
01Share the problem
The product, what is happening, and what you want to change.
-
02Establish the fit
Discuss the goal, access, constraints, and useful next steps.
-
03Agree the scope
Put deliverables, price, timing, and responsibilities in writing.
-
04Work and hand over
Review the result together and leave the team with usable knowledge.
Questions before we start
What kinds of problems are a good fit?
Improving how your engineering team uses AI, from context and planning to implementation and verification. I also help with difficult iOS and macOS problems, including concurrency, audio and transcription reliability, architecture, and delivery.
Can you work with our existing engineering team?
Yes. The scope should fit your team’s ownership, review process, and delivery constraints. The goal is to make progress together and leave the team able to maintain the result.
Do we need a review or implementation?
If the cause or direction is unclear, a focused review is a useful starting point. For AI workflows, we can start with one real task and evaluate it together. If you have an agreed engineering problem and outcome, we can discuss implementation directly.
How would remote collaboration work?
I’m based in Warsaw, Poland. We can discuss time-zone overlap, communication, and access requirements as part of deciding whether the engagement is a fit.
How are price and timing agreed?
After we understand the problem and access requirements, we agree the scope, deliverables, price, and timing in writing. The offers above are starting formats for that conversation.