Anyone can build an app with AI now. Running it safely is the hard part.

Zubl is a platform where AI-generated business applications run as if a platform team had built them: access decided by one engine that fails closed, every change evidenced, data kept in its region, and reporting that needs no BI project. Those are properties of the platform — not things each app has to get right.

In use with design partners now. Looking for design partners and implementation partners in our first markets.

The problem

Generating an app got easy. Owning one didn't.

The hard part was never the screens

Who may see a record, what the audit trail must prove, where the data may live, and what happens when the model changes — every generated app has to answer these, and most answer them badly or not at all.

A hundred apps means a hundred security reviews

When every team can generate its own tools, the organisation inherits a sprawl of small systems with their own logins, permissions and copies of the people list. Nobody can say what is running, let alone prove it is safe.

The enterprise answer has been “no”

So the capability stays in pilots and side projects, and the organisation keeps its spreadsheets. The opportunity is not better generation. It is a place where generated apps are safe to run.

The answer

Apps are definitions. The platform does the rest.

A Zubl app is a definition — records, fields, relationships, lifecycles, rules — not code. The platform generates the screens, the API, the validation and the reports from that definition, and applies the same guarantees to every app, whether we wrote it, you extended it, or the AI builder drafted it this morning.

Access is decided once, for everything

One engine decides who can see and change what, from the organisation structure itself. It fails closed. A new app inherits it on its first day; nobody writes permissions per app.

Evidence is a property, not a feature

Every change is attributed and effective-dated, and reads of sensitive records are logged before the data is returned. An app built this morning is as auditable as one we shipped.

Data stays in its region

Each tenancy is its own database in the region you choose. Residency is a decision made once at the organisation level, not a setting re-discovered per app.

Questions need no BI project

The definition of an app is also its semantic model. Ask a question across any register and the platform already knows what the fields mean and who may see them.

Upgrades do not cost you your changes

Your changes are kept apart from ours. When the platform or an app upgrades, your configuration replays on top — reviewed in advance, never silently overwritten.

AI builder

Describe the app. Review the change. Publish when you're ready.

The AI builder works inside the platform, as the person using it, with exactly their permissions. It drafts an app or a change as a reviewable change set against the same compiler that checks everything we ship. Nothing it writes reaches users until a person publishes it — and nothing it writes can escape the access engine, the audit trail or the residency rules, because those are not the app's to decide.

How the AI builder works →

Ready-made apps

Start with apps that already exist. Extend them. Build beside them.

We ship app families for governance, risk and compliance; legal and contracts; strategy and delivery; operations; finance; people; and community services. All of them share one model of your organisation, and all of them are open to extension. Details, and the full roster, live on the apps site.

Why this, why now

The value moved from writing software to governing it.

For thirty years the cost of a business application was the cost of building it. That cost is collapsing. What remains — and what grows with every app generated — is the cost of running it responsibly: deciding access, keeping evidence, honouring residency, surviving upgrades, answering questions across it.

Zubl is built for that world. The platform carries the guarantees; apps are cheap, many and safe. That is why we lead with the platform, ship the apps as proof, and invite partners to build on it rather than around it.

Where we are, honestly →

Hosted where your data belongs

Dedicated environments in the region you name. The current list is published, with what is live today and what is stood up on request.

Published pricing

Two plans, every app included, in US dollars. If it changes, it changes on the page.

Security answers up front

What we have and what we don't, stated plainly — including what is not yet certified.

You can leave

Full export in open formats at any time, with no exit fee. It's in the terms.

We're working with design partners now.

A small number of organisations shaping the first release, in exchange for early access, direct influence over what ships next, and pricing that reflects the risk of going first.