Frequently Asked Questions

Stack9 is a composable application platform from April9 Growth Tech. It pairs Stack9 Core, a configuration-driven backend and UI engine, with Stack9 Experience, a digital experience platform for content, marketing, and AI.

Stack9 Core is for building custom applications, back-office systems, and portals from JSON entity definitions. Stack9 Experience adds pre-integrated content management, marketing automation, and AI capabilities on top, accessible via API.

Stack9 is a commercial platform built and maintained by April9. Licensing is fixed per instance rather than per seat or per user, so cost does not grow as your member, employer or staff numbers grow. AWS infrastructure is on-charged at actual usage, email is usage-based, and support runs through a Base Support Package with Development Credits for enhancements. See the pricing page for the full structure, or contact us for a quote against your instance count, environments and integration scope.

Install the Stack9 CLI, scaffold a new instance, and follow the getting started guide to define your first entities and screens. Most teams have a working application within a day.

Stack9 is TypeScript end to end: an Express and Knex backend with a PostgreSQL database, a React-based UI, and Redis-backed background jobs, packaged as a set of npm libraries and a base Docker image.

Stack9 Core ships as a Docker image, and four deployment models are supported: cloud-native on AWS ECS in an account you own with April9 operating it (recommended and fully optimised); the same model described as PaaS for procurement frameworks that ask for it; hybrid or self-hosted, where Azure is technically supported with some performance trade-offs; and multi-tenant SaaS where cost efficiency matters more than full infrastructure isolation. AWS is the primary and recommended environment — deployment outside AWS or Azure is not recommended without further validation.

Support is delivered by a fully onshore, Brisbane-based April9 team by phone, web and email, 7.00am to 5.00pm Queensland time on weekdays. The Support Portal at support.april9.au takes tickets 24/7. The standard commitments are at least 99.90% monthly availability including scheduled maintenance, a recovery point objective of 10 minutes and a recovery time objective of 24 hours. At least 80% of incidents are restored within the resolution time for their severity, at least 80% of calls are answered within 20 seconds, and at least 80% of tickets get a response within 2 hours. Level 3 support comes from the April9 engineers who build the platform. Every release carries a 90-day warranty.

Everything stays inside Australian AWS regions: ap-southeast-2 (Sydney) for production, metadata, logs and AI processing, and ap-southeast-4 (Melbourne) for backups and disaster recovery replication. Support is delivered by a fully onshore Australian team with no offshore tier. AI processing residency is worth asking any vendor about specifically — it is where most global platforms have a gap.

No — April9 is not SOC 2 certified and is not pursuing it at this time. What is held is ISO 27001, for over five years, underpinned by a mature ISMS with segregation of duties, least-privilege access and formal risk management. AWS, the underlying infrastructure provider, is itself ISO 27001 certified. For Australian government risk specifically, the more relevant credential is the Comcover engagement, delivered and operating under IRAP certification. The architecture has also been independently reviewed by a third-party security reviewer and by AWS solution architects.

Content and pages, branding and navigation across 50+ Site Settings fields, forms and their routing, email and SMS notifications and templates, customer journeys, entities and screens and automations in Stack9 Console, AI agents against already-connected data, and post-launch adjustment through feature flags. Engineering is needed for genuinely new building blocks: a new UI component, a new webhook target, a new entity type, the first connection of a new data source, and deliberate personalisation logic. After a block is built once, it joins the library for configurators to reuse. The correct claim is "no developer for routine changes" — not "no developers ever".

Not as a toggle, and we would rather say so. Genuine personalisation almost always needs custom logic at the presentation layer, because the parameters — role, organisation, policy type, claim type, relationship — are specific to each organisation’s data model and business rules. Any vendor claiming otherwise is usually describing showing or hiding a banner. What Stack9 does is make that work fast and low-risk: User Groups, permissions and domain entities already exist as configured platform data and are exposed directly to the Web App Template layer, so the logic on top is materially quicker to build than the equivalent against a generic platform. It still requires deliberate development.

Partially, as standard. Accessibility best practice — touch target sizing, semantic labelling, descriptive link text, alt text, colour contrast — is built into the standard design process for client-facing portals, and April9 has delivered a production system specified and approved to WCAG AA. A formal WCAG 2.2 AA conformance audit and remediation plan has also been completed for a client. But dedicated WCAG 2.2 AA compliance testing and independent third-party auditing are not standard inclusions on every engagement; they are scoped explicitly where required. The standard applies to the client-facing experience layer — Stack9 Core’s back-office administration screens are a distinct surface not held to the same bar.

On a managed engagement you do not run upgrades yourself: April9 rebuilds and deploys both the Core base image and your instance image, and every rollout goes through formal Request for Change management. You can stay on an LTS version for as long as its underlying dependencies still receive security patches — that is the honest limit, LTS is not indefinite — with an annual framework uplift recommended and included under the MSA. Mechanically, instances update by pulling new versions of the published npm packages and base Docker image, and bootstrap runs are idempotent so migrations and schema syncs apply safely on restart.

No, by design. No model is trained or fine-tuned on client data. A fine-tuned model absorbs content into weights that cannot be cited, cannot be version-stamped against a change to the source, and cannot be corrected without retraining. Stack9 AI Agents read your authoritative sources when a question is asked, through governed tools and a version-stamped index, and every answer cites the source and version it relied on. Your records stay in your own isolated AWS environment in Australia and are never pooled with another client’s data.

Stack9 AI Studio is the native module where your AI Agent configurators operate agents over their lifecycle. It lets them tune each agent’s instructions, tools, guardrails and model without code. They can replay any conversation step by step, including every tool call and the identity it ran under, and compare versions on accuracy, cost and performance. A version can be promoted or rolled back in seconds. Every promotion must first pass an evaluation set that you approve in advance, and every change goes through change control. Connecting a new data source or tool for the first time still needs a developer. After that, business users can build and run agents against it themselves.

Your right to keep running the solution does not depend on April9. You hold a perpetual licence to Stack9 at its latest running version. The solution runs in a dedicated AWS account provisioned for you, and that account can be transferred to you with root access at any point. On transfer or exit you receive the complete operating package: source code built for you, infrastructure-as-code, CI/CD pipelines, configuration documentation, architecture diagrams and runbooks. Your data can be exported in open formats at any time. This is not a one-click migration to another platform. It gives you the running solution and everything needed for you or another provider to operate it.