SharonAI posts $US294,014 revenue as AI contract claims top $US2.2bn

SharonAI Holdings, the Australian neocloud operator trading as SHAZ, reported $US294,014 in revenue for its first post-IPO quarter alongside a net loss of $US19.9 million, even as it pointed investors toward more than $US2.2 billion in disclosed AI infrastructure contracts. The March-quarter filing draws a clear split between booked turnover and headline deal value. This remains a capital-intensive build-out story, not a software earnings one.

In the same 10-Q summary, SharonAI disclosed that one customer agreement with ESDS was worth $US1.26 billion over 60 months. A separate company statement carried by Business Wire and The AI Journal said the group had also disclosed a further $US950 million customer contract on 13 May. Those figures sit next to a quarter that produced less than $US300,000 in revenue. They explain why the stock is being read less as a conventional earnings play and more as a financing and execution test for a young AI compute operator.

Contracts and revenue are different things. SharonAI’s filing makes that gap hard to miss: the headline deal values sit beside a top line that is still barely off the ground.

The balance sheet tells the same story. SharonAI’s filing put the fair value of its convertible notes at $US199.4 million at 31 March, a figure that dwarfs the quarter’s revenue base. Early-stage infrastructure groups routinely spend ahead of revenue, but SharonAI’s numbers show how far this business remains from a steady-state operating profile. Investors are being asked to weigh long-dated contract announcements against a quarter that is still very small in accounting terms.

Management stayed focused on growth, not near-term income. Chairman, chief executive and co-founder James Manning said in an earnings call transcript that SharonAI was “well placed to continue to grow our customer base over the coming quarters”. The statement fits the broader message in the filing: SharonAI wants investors to watch customer wins, contracted capacity and pipeline scale, even though recognised revenue is still modest.

The company is pitching a domestic AI infrastructure story at a time when enterprises are hunting for compute capacity without building it themselves. SharonAI has placed itself inside that neocloud conversation. But the quarter suggests the harder part starts after the announcements. The company now has to show that multi-year contract values translate into billable services, delivered capacity and a revenue run rate that begins to match the headline figures.

The 60-month ESDS agreement is a case in point. A contract of that size may support the long-term argument, but its economics will only become visible over time, not in one quarter of reported sales. The same applies to the separate $US950 million deal disclosed in May. Disclosed contract value is not revenue recognised in the period, and it is not proof that the build-out has become self-funding. Until those agreements begin feeding through into recognised revenue, margin visibility and cash generation, the filing leaves SharonAI looking more like a company financing AI infrastructure growth than one already collecting at scale from it.

The contracts still matter. They are the reason the quarter deserves attention. But the March filing sets a harder baseline for how SharonAI should be judged. It is no longer enough to cite demand for AI compute in the abstract. The next few quarters will need to show whether customers are consuming services on the timetable implied by the announcements, and whether the company’s funding model can support that build-out without further strain.

SharonAI has shown it can announce large deals in a market still chasing AI infrastructure exposure. What it has not yet shown, on the numbers filed for the March quarter, is how quickly those deals turn into the kind of recurring revenue that can carry a public company through the expensive early years of building compute capacity.