KIDZ AI is a live after-school tutoring platform that spent the second quarter renaming itself, converting most of its convertible notes into stock, and announcing a compute-leasing strategy that still produces no revenue. The investment debate is whether a sub-scale classroom business can become a contracted graphics-processor landlord before dilution and listing pressure consume the residual claim. Cash on the mid-year balance sheet looks large next to the tutoring franchise. That cash arrived through equity sales and a crypto unwind rather than from operations, which is the distinction the market is still pricing poorly.
The load-bearing event after the quarter closed is the Canopy Wave compute-services agreement. Catalyst Compute, a wholly owned subsidiary, is supposed to order a dedicated cluster of NVIDIA HGX B300 accelerators and lease that capacity for five years. The stated contract value is about $45 million, but the agreement is conditioned on a non-cancellable server order that the mid-year accounts do not show as placed. Until that order is funded and the cluster is live, the neocloud story is a signed letter sitting on top of a shrinking subscription book. Classroom revenue still pays the teachers. It does not pay for a Blackwell deployment.
The same quarter that management called an inflection also cut tutoring demand. Service revenue printed $482 thousand as selling spend was halved and attention moved to compliance, treasury, and the rebrand. Gross margin held near the mid-forties, which shows the classroom unit still works when a student pays. The problem is scale. Operating expenses still dwarf that gross profit, and the quarterly report itself raises substantial doubt about continuing as a going concern. Whether Catalyst Compute funds the graphics-processor order, and whether any compute revenue appears before the next listing review, decides if this is a platform transition or another special-purpose-acquisition narrative cycle.