Software · Mobile · AI
Software and AI products, built end to end.
We build the app, the backend and the model that makes it useful. One senior team takes your product from first prototype to the App Store.
Our work
Products we design, build and ship.
Mobile apps, web platforms and the maths and models inside them, taken from first prototype to launch.
-
2026
Ham radio · iOS, Android + Web
SSTV Studio
An app that decodes slow-scan TV pictures live from a radio and turns photos into SSTV audio, with an ISS pass tracker that points your antenna using the phone’s sensors.
- Delivered
-
- Real-time decoder and encoder, 14 modes
- ISS tracker with Doppler and sensor pointing
- iOS and Android apps from one codebase
- Stack
- JavaScript · Web Audio · Capacitor · three.js
-
2026
Ham radio · Web, iOS + Android
Antenna Lab
An interactive simulator for ham radio antennas. Change the geometry, balun and feed cable and the 3D pattern, impedance and SWR sweep update as you move each control.
- Delivered
-
- 3D simulator for five antenna families
- Method-of-moments Yagi solver checked against NEC-2
- Offline web app plus iOS and Android builds
- Stack
- TypeScript · three.js · Vite · Capacitor
-
2026
Fintech · Mobile + AI
Example projectIn-app support assistant for a personal finance app
A retrieval-based assistant that answers account and billing questions inside the iOS and Android apps, cites its sources, and hands off to a human when it is unsure.
- Delivered
-
- Assistant UI in both native apps
- Retrieval pipeline over help content
- Evaluation suite and escalation rules
- Stack
- Swift · Kotlin · Python · PostgreSQL
-
2025
Operations · Mobile + on-device AI
Example projectOffline photo inspection app for field teams
A mobile app that flags issues in site photos with an on-device vision model, works without signal, and syncs reports when the connection returns.
- Delivered
-
- React Native app
- On-device model and update pipeline
- Sync service and admin dashboard
- Stack
- React Native · Core ML · TensorFlow Lite · Node.js
-
2025
Health · Web + AI
Example projectPatient intake and scheduling platform
A booking platform where patients describe their symptoms in their own words and a language model routes each request to the right clinician and appointment length.
- Delivered
-
- Patient web app
- Clinic dashboard
- Triage model with clinician review
- Stack
- TypeScript · Next.js · Python · Firebase
Process
From first call to launch in 10 to 20 weeks.
Five steps, one team, and a working build in your hands every week. You always know what ships next and what it costs.
- 1–2 weeks
Discovery
We map the product, the people who will use it and the riskiest assumption in the plan. You leave with a clear scope and a fixed estimate.
- Scope and priorities
- Architecture sketch
- Fixed estimate
- 2–4 weeks
Prototype
A working slice in your hands: a real app with a real model behind it. We put it in front of users before anyone commits to the full build.
- Working prototype
- User test findings
- Model feasibility report
- 6–12 weeks
Build
Two-week cycles with a demo every Friday. The app, backend and AI ship together from one team, on one schedule.
- Weekly demo builds
- Evaluation suite for the AI
- Shared progress board
- 1–2 weeks
Launch
App Store and Google Play review, monitoring, and checks that the model behaves in production the way it did in testing.
- Store releases
- Monitoring and alerts
- Launch checklist
- Ongoing
Handover
Documented code, clean repositories and a runbook your future team can own. Keep us on for support, or take it from here.
- Documentation and runbook
- Repository transfer
- Optional support retainer
Contact
Tell us what you're building.
Send a few lines about the product, your timeline and where you're stuck. An engineer, not a salesperson, replies within one business day.
contact@makex.studio
Typical engagements start within two to three weeks of a first call.