Our Story

We watched brokers drown in paperwork.

Insurance Advice was born from a simple observation: the smartest people in insurance spend their worst hours reading PDFs instead of advising clients.

The problem

Eighty pages. Every single time.

An Australian insurance broker reviewing a QBE Business Pack sits down with an 87-page PDS, a highlighter, and a spreadsheet. They cross-reference exclusions against the schedule, check sub-limits against the client's assets, and hunt for conditions that could void cover at claim time.

That process takes 2–4 hours per policy. For a renewal with three competing quotes, it's a full day of reading before the broker can even begin advising.

And when they're done? There's no audit trail. No structured evidence for ASIC's Best Interests Duty. Just a file note and a prayer that they didn't miss the sub-limit on page 67.

We thought: what if the reading took 60 seconds?

The origin

Built in Sydney. For Australian insurance.

Insurance Advice was founded in 2025 in Sydney. We'd seen the gap firsthand — brokers, underwriters, and risk managers all spending their most valuable hours on manual document review instead of the professional judgement they were trained for.

The insurance industry doesn't have a talent problem. It has a time problem. The most experienced broker in your firm and the newest recruit both have to read the same 80-page PDS. The difference is what they do with it afterward.

We asked: what if AI could handle the reading — the extraction, the cross-referencing, the gap detection — and give the professional back the hours to actually advise?

Our approach

Subscription-only. No commissions. Ever.

Insurance Advice is a subscription SaaS tool. We don't take commissions. We don't have insurer partnerships that influence our output. We don't earn more when you place with one insurer over another.

Our AI analyses every policy the same way — coverage by coverage, exclusion by exclusion, condition by condition. The output is structurally identical regardless of which insurer issued the document.

This matters because the biggest unspoken problem in insurance technology is vendor bias. When your analysis tool is funded by insurers, it's hard to trust the output. We're funded by you — the professionals who use the platform.

Independence isn't a feature. It's the foundation.

The journey

Where we've been. Where we're going.

Late 2025
The idea. We identified the PDS review bottleneck. Research began into AI-driven policy analysis for Australian insurance.
January 2026
First prototype. Claude analyses a real QBE Business Pack PDS — 48 pages, full structured output, every exclusion cited. 58 seconds.
February 2026
Platform build begins. Database architecture, API design, and multi-tenant isolation — building properly from the foundation up.
March 2026
End-to-end pipeline live. Upload a PDF → Claude analyses → structured JSON → results displayed. Stripe billing working. Auth flows complete.
April 2026
Ask IA and dashboard. Natural language chat with uploaded policies. Firm dashboard with broker activity tracking. Marketing site fully built out.
July 2026
Public launch. Fully functional platform live and accepting paid subscriptions. Coverage Comparison, PDS Library, and client-ready exports.
Our values

What we believe. What we build around.

Accuracy over speed

A confident wrong answer is worse than no answer. We'd rather flag uncertainty than fabricate confidence.

Independence is non-negotiable

No commissions, no referral fees, no insurer funding. Our only customer is the professional using the platform.

Build properly, not quickly

One feature that works perfectly beats five that need patches. Architecture first, polish second, shortcuts never.

AI assists, professionals decide

The AI reads, extracts, and flags. The licensed professional reviews, decides, and advises. That line never blurs.

Transparency by default

Every AI output explainable. Every citation verifiable. Every confidence score has a rationale.

Your data is yours

Australian-hosted, tenant-isolated, encrypted. Anthropic doesn't train on your data. We don't sell it.

Contact

AI-powered insurance.
Coming soon.

We're here to answer your questions. We usually reply within one business day.

Prefer email? Write to hello@insuranceadvice.app