Built by a safety practitioner for the SMEs no consultant calls back.
I'm Aaron. I have a postgraduate qualification in occupational health, safety and environment, and I've spent enough time inside Australian SMEs to know one thing: the businesses that need a real safety system the most are the ones who can least afford the consultancy hours to build it.
RAE IQ is the platform I wish I'd had when I was the safety lead. A copilot that drafts the documents, runs the registers, tracks the obligations and grounds every output in something you can defend, without paying $300 an hour.
Built in Australia, for Australian work.
Sources a court would accept.
Every regulatory output the platform produces is traceable to a defined 7-tier authority hierarchy. Top to bottom, what we cite and why we cite it.
| Tier 1 | Binding law | WHS Acts, regulations, mandatory codes |
| Tier 2 | ISO standards | ISO 45001, 45003, 19011 and the family |
| Tier 3 | Regulator guidance | Safe Work Australia, state regulators |
| Tier 4 | Adjacent Commonwealth law | Privacy Act, Fair Work Act, NHVR |
| Tier 5 | Industry regimes | NHVR HVNL, mining safety, electrical safety |
| Tier 6 | Professional bodies | AIHS, RACP, ACOEM, AASCN |
| Tier 7 | Insurer & union guidance | WorkCover, ReturnToWorkSA, AWU, CFMEU |
The full reference is published in our repository at docs/AUTHORITATIVE_SOURCES.md, with a live audit at docs/AUTHORITATIVE_SOURCES_AUDIT.md showing which sources back which features. We don't know any other compliance platform that publishes this.
Two national transitions we're already tracking
Built for every Australian jurisdiction.
Terminology, legislation citations and regulator names adapt to your state automatically. Pick Queensland and your SWMS references Workplace Health and Safety Queensland; pick Victoria and it references WorkSafe Victoria. PCBU vs employer is handled for you.
How the copilot works.
The honest explanation of what's happening behind every drafted document, in one paragraph.
The drafting engine uses Anthropic's Claude model with structured-output schemas (Zod) so the response shape is verified at generation time. Jurisdiction context is prompt-cached so we only pay to read it once. Your reference documents (uploaded SOPs, policies, client requirements) are injected into the prompt for that single generation, and dropped immediately afterwards.
Customer data is never used to train AI models. Prompts are short-lived and isolated to your organisation; nothing you upload leaves the platform's tenancy.
Security posture
Every table is protected by row-level security; every organisation is fully isolated; everything is encrypted in transit and at rest.
- Row-level security on every table: no cross-organisation data leaks possible.
- Encrypted in transit (TLS 1.2+) and at rest (Postgres + Supabase Storage).
- Audit logs on medical record access and other sensitive workflows.
- Privacy Act 1988-aligned defaults for worker health monitoring data.