SCX.ai seeks AUD $40M in ASX float for local AI inference
What's the deal? SCX.ai has launched a fully underwritten initial public offering to raise AUD $40 million and list on the Australian Securities Exchange under the code SCX. The AI infrastructure provider is offering 133,333,334 ordinary shares at AUD $0.30 each, with HenslowDealroom has a profile for this one. Try Dealroom → and Canaccord GenuityDealroom has a profile for this one. Try Dealroom → acting as joint lead managers and underwriters.
What's the endgame? The company focuses on AI inference — the stage where trained models generate outputs in response to user or software requests. Its platform combines inference hardware, data-centre infrastructure, private interconnection, and management software, positioned around Australian-operated infrastructure for organisations seeking local data residency and governance.
Why now? SCX.ai points to rapid expected growth in inference spending. A report cited in its prospectus estimated Australia's AI inferencing-as-a-service market would grow from about AUD $0.3 billion in 2025 to AUD $3.5 billion by 2030, with the global market rising from about $15 billion to $177.4 billion over the same span.
By the numbers: As of the end of May, SCX.ai had AUD $5.4 million of contracted annual recurring revenue and AUD $14.6 million of unbilled total contract value. Its platform had 298 active users as of late June, including paying users and those in pre-subscription trials — a sign the rollout remains early stage.
Proceeds will fund AI hardware and support services, customer acquisition, working capital, and transaction costs. Its initial operational node runs at Equinix SY5 in Alexandria, Sydney, using inference systems from SambaNova Systems.
What could go wrong? Investors are scrutinising how AI infrastructure firms expand without the heavy energy, cooling, and property demands of conventional graphics processing unit clusters. SCX.ai argues its inference systems deploy within existing commercial data-centre footprints rather than purpose-built facilities.
"AI is moving rapidly from experimentation into production," said David Keane, adding that the conventional path for AI inference at scale "requires dedicated facilities built around the power density and cooling demands of GPU clusters."
The signal: The float reflects a broader push in Australia for sovereign technology infrastructure, particularly among government agencies and regulated industries that want data to remain onshore and under local control. SCX.ai is betting that demand carries it from AI testing into live enterprise and public-sector deployment.
Read more: IT Brief Australia
Image credit: cbowns
