Discovery and architecture
Scope, user flows, technical architecture and a written product requirements document. The unglamorous phase where most of the money is either saved or lost.
Salt · Build
Web and cross-platform mobile products, taken from architecture to the app stores and kept alive afterwards.
The failure is rarely the launch build. It is the codebase nobody outside the original developer can read, the store rejection nobody planned for, the absence of analytics that makes every roadmap argument a matter of opinion, and the moment the freelancer stops answering.
We build assuming someone else will maintain it — including you. That changes the architecture, the documentation and what we hand over.
Scope, user flows, technical architecture and a written product requirements document. The unglamorous phase where most of the money is either saved or lost.
UX and UI built to Apple's Human Interface Guidelines and Material 3 so the app feels native on each platform, reviewed for accessibility rather than retrofitted for it, and delivered as a design system your team can extend.
Web applications and cross-platform mobile in React Native or Flutter, with native work where the hardware demands it. Two-week sprints, a working build at the end of each one, and no phase where you cannot see the product.
Automated and manual testing across the device matrix, then submission to the App Store and Google Play handled by us — including privacy manifests, data safety declarations and the review cycle.
Event tracking wired to the decisions you need to make, crash and performance monitoring, over-the-air updates, and an agreed support arrangement so the app has an owner on Monday morning.
Users, flows, scope, architecture, and the one metric the first release has to move.
Design and engineering sprints, shipping to TestFlight and Play internal testing from early on.
QA across devices, performance, accessibility and store compliance ahead of submission.
Release, watch the analytics, and iterate against what real users do rather than what the roadmap assumed.
Cross-platform for most products: one codebase, two stores, meaningfully lower cost to maintain. Native when the product depends on hardware, heavy graphics or platform features that cross-platform frameworks reach late. We decide in discovery, not by default.
Yes, including privacy manifests, data safety forms and the review cycle. Store rejection is a scheduling risk, so we plan for it rather than discover it.
That is a normal outcome and we build for it. Code, pipelines and accounts are yours throughout, and the handover includes documentation and a working session with your team.
Describe what it needs to do and who it is for. If a smaller first release would get you the same answer sooner, we will say so.