Our own products
We build our own apps, not just other people's.
Client work pays the bills; our own products keep us honest. Everything we ask a client to trust us with — architecture, store compliance, AI guardrails, release engineering — we do first on software with our own name on it.
Remedy Lab
Ask a question in plain language. Get clear, step-by-step instructions for natural remedies and homemade products — with hazard screening on every single response.
Remedy Lab is an AI-native mobile application built with .NET MAUI. The distinguishing feature isn't the AI; it's what surrounds it. Every response passes through automated hazard screening and explicit refusal handling before a user ever sees it, because "mostly safe instructions" is not a standard anyone should ship.
Remedy Lab
"Something soothing for a sore throat?"
Instructions
Not for children under 1 year — honey carries a botulism risk in infants.
Under the hood
Why this app is our best sales pitch.
Remedy Lab is a working demonstration of how we think AI products should be engineered. Everything below is shipping, not aspirational.
The AI runs on your phone
By default, inference happens on the device itself. No round trip to a server, no per-query cost, and no user data leaving the handset. The app works in full with no signal at all — which for a product people reach for in a kitchen or a cabin is the point.
Guardrails in every response
Automated hazard screening and explicit refusal handling are part of the response pipeline, not a disclaimer in the settings screen. When the safe answer is "don't do this," the app says so plainly instead of improvising.
One schema, three engines
A single response contract is served interchangeably by an on-device model, a model hosted on a local network, or a cloud provider. Swapping engines is configuration, not a rewrite — the same architecture we build into client systems.
Held to a hard engineering bar
Layered architecture enforced by automated architecture tests, a standing 90% code-coverage gate, zero-warning builds, and full CI/CD through signed, store-ready release packaging. The build fails before a bad design ships.
Built with
Platforms
- Android Available now on Google Play.
- Windows Built from the same codebase.
- iOS In development.
What this means for you
Nothing we propose is theoretical.
When we recommend on-device inference, a provider-agnostic layer, or a coverage gate, we're describing something we already operate — including the parts that were harder than expected.
It also means our estimates account for the whole job. Store review cycles, signing certificates, subscription restore behavior, and privacy declarations are in the plan from the start, because we've paid for underestimating them ourselves.
More on the way
Our product line is growing.
We maintain Remedy Lab actively and have further applications in development. If you'd like to hear when something new ships — or want to talk about building your own — get in touch.
Next step
Want an app built to this standard?
Tell us what it should do and who it's for. We'll come back with a realistic scope and an honest read on what it takes to keep it alive after launch.