Diamond EdgeCutting edge. Rock solid.
Kotlin Multiplatform consulting
I'm Reed Ellsworth, and Diamond Edge is my independent practice. With AI doing the routine work, one senior engineer now delivers in weeks what used to take a team a quarter. I build native Android and iOS apps from one Kotlin Multiplatform codebase, at a fixed fee that's a fraction of a conventional team's, with the source in your repository from the first commit.
Three apps of my own, built the way I'd build yours: one codebase, tested on real phones, submitted to both stores. Each case study is less about the app than about what the same work would look like on your project.
On-device AI · Offline
Dictation for lawyers and financial advisors that works with every radio off. Speech recognition, decoding and the finished documents all happen on the phone; the Android build has no network permission at all.
For you: a feature that has to work without a connection, or data that can't leave the device.
Real-time · Server
Eight card and board games with live multiplayer across phones, a server I built and run, and game rules that update without an app release.
For you: connections that drop and recover, one device as the source of truth, and a backend written in the same language as the app.
Design system · Release
A score pad for card nights with a hand-drawn look, built on a theme library I wrote and open-sourced. It collects no data at all.
For you: a small app with its own brand, finished properly and shipped to both stores.
Most of the cost of a conventional mobile project was never the hard engineering. It was the volume of routine code, written twice by two teams, and the coordination between them.
The routine work: data layers, boilerplate, tests, mechanical conversion of old code, store paperwork. What used to take weeks now takes hours.
Everything that decides whether the app works: architecture, integration with your systems and devices, measuring on real phones, store review, and saying no to the feature that isn't ready. AI doesn't make those calls; I do, and I answer for them.
Fast only helps if the code holds up. It's in your repository from the first commit, so your engineers can read every line, and review it before you sign off.
The result is projects that never cleared the budget now can: the customer app you priced once and shelved, the companion app for a product line too small to justify a mobile team, the migration you've deferred for three years, the client app an agency used to turn down.
Any app for Android and iOS: customer-facing or internal, consumer or business, or the mobile side of your web product, with the backend it needs. Most projects arrive in one of these situations. If yours is a poor fit, I'll tell you so and point you somewhere better.
Any company
"We know what the app should do. We don't have a mobile team."
The app, the backend it needs, and App Store and Google Play submission under your accounts. The three apps above show the range, from AI that runs on the phone to real-time multiplayer with its own server.
Hardware companies
"The hardware ships in ten weeks and we have no app team."
Pairing, connection management, device control, and the screens your customers use every day. Your team owns the device; I build the app that talks to it, with its side of the device protocol written once for both platforms.
Agencies
"A client asked for an app and we have nobody to build it."
I work under your brand, never contact your client, and hand back code you can present as your own. No direct contact and no solicitation, in writing.
Teams on dying frameworks
"Our app runs on Xamarin, and Microsoft stopped supporting it."
Xamarin, Cordova, Ionic and aging React Native, ported to Kotlin Multiplatform with the tests carried across, and the mechanical layers converted by tooling rather than retyped.
One fee, agreed before work starts, never an hourly rate. You're buying an outcome, often against a date you've already committed to, so that's what the price is set against. If I've scoped it wrong, the overrun is mine.
Published, because a fraction of the usual cost should be something you can check. Your actual fee comes out of the assessment, in writing, before you commit to anything more.
One to two weeks of real work at a fixed fee, credited in full toward the engagement if you go ahead. If you don't, you keep everything it produced.
| Assessment | Fee | What you get |
|---|---|---|
| Protocol Sprint, for hardware | $5,000 | A review of how your device talks to an app, a written interface contract between the two, a working connection prototype against your real device, and a fixed quote |
| Migration Assessment | $4,500 | A codebase audit, a complete feature inventory, a dependency and risk map, and an effort model you can take to your own budget process |
| Scoping, for a new app or an agency brief | $2,500 | A review of what the app needs to do, an architecture outline, and a fixed quote; for agencies, one you can price your proposal against |
| Engagement | Most land between | Typical length | The conventional route |
|---|---|---|---|
| A new app, including device companion apps | $20k – $80k | 3–8 weeks | An agency quote of $180k–$400k and 6–9 months |
| Migration to Kotlin Multiplatform | $30k – $100k | 4–12 weeks | Two native rewrites, one per platform, by two teams |
| Agency project, under your brand | By conversation | 3–8 weeks | Turning the work away |
Me, not a team assembled around your project. I scope the work, write the code, handle store submission and answer for the date.
Before Diamond Edge's current apps I spent 14 years building Android: the sole Android engineer on Weber's Summit Smart Valve grill, with live temperature and cook state streamed over BLE and Wi-Fi; Android lead for a point-of-care kiosk fleet of more than 100,000 devices at Outcome Health; and principal engineer on Sling TV's live-streaming app, with 1M monthly Android users. I also wrote KmLogging and DiamondCharts, two open-source Kotlin libraries on Maven Central.
Thirty minutes, technical, no deck. Bring your launch date, your device's protocol documentation, your existing codebase or your client's brief. You'll leave with a straight read on scope, fee, and whether I'm the right person for it.