Work with me
One person, from the idea to the App Store.
Most product work loses more time to the handoffs than to the building. I do the design and the engineering, so there is no handoff — which usually means the thing exists weeks earlier and looks like one person decided it, because one person did.
Arrit Gashi
Product Engineer · Huntersville, North Carolina
What I take on
Ship a product, start to finish
From an idea to something people can download. Design, build, store submission, and the review correspondence nobody warns you about.
- Product definition and the design that goes with it
- A working app — iOS, Android, or web
- Backend, auth, and data model
- App Store and Play Store submission, including review responses
- The source, documented well enough to hand to someone else
AI features that survive contact with users
Most AI features fail on the cases where the model is subtly wrong rather than obviously wrong. Designing for that is the job.
- Product flows built around real model output, not a happy path
- Guards against confidently wrong answers
- Model selection matched to task and cost
- Evaluation you can run again after a prompt or model change
Design systems that hold up in production
A component library that engineers actually adopt, because it matches the code they already write and it is cheaper to use than to bypass.
- Tokens, components, and the rules for using them
- Accessibility verified by measured contrast, not by eye
- Documentation written for the people who have to implement it
- Automation for the repetitive parts of production design
Prototype to production
You have a prototype, a Figma file, or a proof of concept, and it needs to become software that can be deployed and maintained.
- An honest assessment of what carries over and what has to be rebuilt
- A production architecture with the tradeoffs written down
- The build itself
- Deployment, and a path for whoever maintains it next
How it runs
- 01
A call, then a written scope
Thirty minutes to understand the problem. You get back a written scope with what I would build, what I would not, and what it costs. No charge for this part.
- 02
A working thing in week one
Not a deck and not a wireframe — something that runs. It will be incomplete, and it will tell you more about the decision than a month of documents would.
- 03
Weekly builds you can open
Every week you get a build and a short note on what changed and what is still open. If the direction is wrong, week two is a cheap place to find out.
- 04
Handover that actually hands over
Source, deployment, credentials, and documentation written for whoever comes next. The decision records explain why things are the way they are, not just what they are.
Start a project
Email is the fastest route. Tell me what you are trying to build, who it is for, and when you need it — a paragraph is plenty. If it is not something I should take on, I will say so and point you somewhere better.
Based in Huntersville, North Carolina. Remote work across US time zones, and European mornings.