We build iOS, Android and web apps, AI chatbots, voice assistants and workflow automations.
Services
Across apps, automation and AI features
8
Apps built in-house
Both are live on the App Store.
2
Ads and trackers
In anything we ship. Client data isn't used to train AI models.
0
Tools we work with
Apps
Software people use every day.
Automation
Fewer manual steps for your team.
AI features
Assistants and vision, inside your product.
The honest GLP-1 companion.
Track your GLP-1 medication doses, weight and protein in one calm place, with a free tier that stays free.
Radar you can read. Every storm since 1991, bought once.
Pro-grade weather radar made legible. Scrub every storm back to 1991, watch live radar, and get a plain-language narrator. A one-time purchase.
A short call, then a written brief: the problem, who it's for, what v1 includes and what it will cost. For automation work it also estimates the hours saved. The brief is yours to keep, whoever builds it.
Key screens for apps and web products. Scripts and conversation flows for chatbots and voice agents. You approve the design before we build, which keeps surprises for later to a minimum.
Each week you get something you can use: a TestFlight or Play build, a staging link, or an agent you can talk to. Your feedback goes into the next week's work.
Store submission or production deploy, monitoring and alerts, written docs and a walkthrough for your team. Code, accounts and API keys stay in your name.
A weekly call and written updates in between. When you ask for something new, you hear what it costs and when it will land before any work starts.
We keep data on the device where we can and collect as little as possible where we can't. Nothing is sold, and your data isn't used to train third-party AI models.
Some processes aren't worth an AI build. We work out the cost and the time saved first, and if the numbers don't add up, we say so on the first call.
Code, repositories, prompts, agent workflows, API keys, store listings and design files live in your accounts from day one. You can move to another team at any time.
People decide whether they trust software within the first minute. We spend real time on layout, wording and edge cases.
What you won't get
2-week sprints
For ongoing work on an app, an AI feature or a set of automations. We agree the scope at the start of each sprint.
$2,500USD / sprint
Pause or stop between sprints
Book a callWhat's included:
Fixed-scope v1
One product taken from idea to launch: an app on one platform, or one chatbot, agent or automation.
$5,000USD
Scope and price agreed before we start
Start a projectWhat's included:
Larger projects
For apps on both iOS and Android, multi-agent systems, voice and computer vision work, or a monthly retainer.
Custom
Quoted after a scoping call
Get a quoteWhat's included:
Not sure which fits?
Book a free 30-minute intro call and we'll scope it together.
The studio
A software and AI studio working with clients around the world. We build mobile and web apps, AI chatbots and voice assistants, and workflow automations for clients. We also build and publish our own apps, Onward and Backshear.
Apps for iOS, Android and the web. Automations, including AI agents and social media workflows. And AI features such as chatbots, voice assistants and computer vision. The services page covers each one.
Working together
A fixed-scope Idea → Launch project is $5,000 USD. Ongoing work runs in 2-week Build Sprints at $2,500 USD each, and you can pause between sprints. Larger projects are quoted after a scoping call. AI projects also have running costs for model and API usage, which we estimate up front.
Yes. Privacy policy, support and marketing pages, store listings, privacy labels, screenshots and replies to review feedback are all included.
You do, from the first commit. Repositories, prompts, API keys, store listings and design files live in your accounts.
AI projects
Usually models from Anthropic (Claude) or OpenAI, and open-source models where running them yourself makes sense. We choose per task on quality and cost, and keep the model swappable.
We use API providers that don't train on your data by default, keep only what the task needs, and store it in your accounts. For sensitive data, we agree where it is processed before we build anything.
It will sometimes, so we plan for it. Uncertain answers go to a person, risky actions need approval, and we test against your real examples before launch and after every change.
Our apps
Go to the app's support page (linked in the footer) or email us. A real person replies.
