Investor perspective
Australia’s AI success is no longer a one-company anecdote
The strongest evidence is not that more businesses now use the letters “AI”. It is that Australian startups are crossing the thresholds that matter: global customers, institutional growth capital, rising private marks, public-market access and strategic exits.
Australia has always produced strong technical talent. What has changed is the conversion of that talent into a visible AI company pipeline. Firmus has become a multibillion-dollar AI-infrastructure story. Neara crossed the A$1 billion threshold by applying AI and physics to electricity networks. Heidi Health and Harrison.ai are embedding clinical AI into healthcare workflows. Lorikeet and Relevance AI are building the agentic software layer. Canva’s acquisition of Leonardo.Ai supplied something the local ecosystem has historically lacked: a meaningful, realised generative-AI exit.
These companies are not interchangeable. Their economics, risks and milestones differ. But together they show that Australia can create and finance AI businesses across three layers: the infrastructure that supplies compute, applied software that owns a valuable workflow, and scaled platforms that distribute AI to millions of users.
The opportunity is not “Australian AI” as a single trade. It is backing focused companies before product proof becomes obvious—and before global growth capital reprices the opportunity.
The applied-software companies are the seed-stage lesson
Firmus is the largest valuation headline, but it is not the cleanest comparison for a software seed investment. The more relevant pattern is visible in Neara, Heidi, Lorikeet and Relevance AI: begin with a narrow, expensive workflow; achieve repeat usage; make the product harder to remove through data, integration, approvals or distribution; then use global capital to accelerate.
Neara’s A$90 million Series D delivered a reported A$1.1 billion post-money valuation after the company expanded from Australian network utilities to major offshore customers. Heidi moved from an approximately US$100 million early-2025 mark to US$465 million in October 2025 as clinical usage scaled. Lorikeet moved from a reported A$100 million mark in February 2025 to more than A$200 million six months later. Relevance AI’s valuation remains undisclosed, but its US$24 million Series B demonstrates that global investors will finance Australian-founded agentic software when the product story is sufficiently strong.
None of this guarantees the next seed company will repeat the trajectory. It does show the pathway exists—and that Australian location is not a ceiling when the customer problem is global.