The clinical app you wish existed, built and deployed on safe, sovereign infrastructure. Compliance, consent and data residency are built in by default.
Real tools Australian clinicians could stand up on Foundry. Pick one to drop it into the prompt above, or swipe for more, there are plenty.
You describe it; Foundry builds it and runs it through the assurance pipeline before anything reaches a patient. Security, data-leak, prompt-injection and clinical-safety checks run on every deploy. Most findings are fixed for you automatically; anything that needs your call is explained in plain English.
Security scan passed. No exposed secrets.
No identifiable patient data leaving the platform.
Prompt-injection tests passed. 34 attempts blocked.
Consent capture present on patient-facing forms.
Patient identifier spotted in an AI prompt. De-identification applied automatically before the model ever saw it.
Ready for preview. Shared with 0 colleagues.
AI made building the tool the easy part. The hard part is everything around it: keeping patient data onshore, staying compliant, making sure nothing unsafe reaches a patient. On Foundry, that's all handled for you, so you can just build.
Patient data stays onshore, encrypted and fully auditable, so you always know exactly where it lives. No consumer platform, no guesswork.
Privacy Act alignment, consent capture and safety checks are part of the platform, not paperwork you assemble. It protects your patients, and your registration, by default.
No more brilliant app stuck on a laptop because there was nowhere safe to run it. Foundry is the compliant home built for the solo clinician-builder.
You focus on the idea. Foundry takes care of the rest.
The moment your app lands on Foundry, it runs on security foundations that take healthcare companies years to build. None of it needs configuring. None of it can be switched off by accident.
Australian-hosted, end to end. Patient data never leaves the country.
Privacy Act alignment built into the platform, not rebuilt per app.
At rest and in transit, without you writing a line of it.
A complete audit trail, always on, ready when someone asks.
Foundry provisions the database and separates dev, staging and production. The AI that helps you build only ever touches synthetic dev data, never a real record.
Security, data-leak, prompt-injection and clinical-safety checks run automatically. Findings are either fixed for you, or explained in plain English.
Compliance here isn't paperwork you assemble. It's a property of the platform your app runs on. That's the difference between a place that lets you run code, and a place built to protect the patient at the other end of it.
Hardened inside our own clinical products, in real practice. Your app imports them like any other building block, except these have already passed clinical review. Vercel doesn't have them. The model labs never will.
Eclipse & Medicare billing. MBS items, claiming, patient payments and GST, handled correctly from the start.
Scribe-tuned models. The clinical scribing and history-taking engines behind our own products, as building blocks.
De-identification as a service. Applied before AI calls. Nobody can safely vibe-code this alone.
EMR extraction. The way in and out of Australia's walled-garden clinical record systems.
AHPRA verification. Confirm users are registered clinicians before they touch anything clinical.
Patient consent capture, built into the platform rather than bolted onto each app.
Electronic prescribing with eRx conformance, a heavyweight regulated primitive nobody vibe-codes.
Compliant SMS & email. The patient-comms plumbing every app needs and nobody wants to rebuild.
Install tools built by other clinicians, including Akuru's own, and use them in your practice. No building required.
Describe what you need and deploy it safely, or bring an app you've already built. Free to build for yourself. You only pay when it reaches your clinic or patients.
Have a great model or component? Publish it for other builders at a rate you set. Foundry handles compliance, billing and distribution.
The coding tool is your choice. Foundry is where it lands: managed data, provisioned environments, primitives over REST and MCP, and a pipeline that gates every promotion to production.
Built it in Claude Code, Cursor, or anything else? Connect it and Foundry rebuilds and deploys it on compliant infrastructure, same pipeline, same guarantees. foundry deploy
Every primitive is an API and an MCP tool. claude mcp add foundry and your agent can call billing, de-id, scribe and EMR extraction directly.
The coding agent only ever reaches the dev database, seeded with realistic synthetic patients. Promotion to production goes through the pipeline, never the agent's hands.
Point at Anthropic, OpenAI, Gemini or Akuru's onshore models through a single gateway, with prompt-safety layers applied on the way through.
Same idea as up top: tell us what you'd build, then leave your details. We'll reach out as soon as your cohort opens.