Note 6. Accounts Receivable, Net

 ACCOUNTS RECEIVABLE, NET

Accounts receivable represents amounts billed and unbilled for services provided under usage-based arrangements. Unbilled accounts receivable represents services performed but not yet invoiced, for which the Company has an unconditional right to payment; accordingly, unbilled amounts are classified as accounts receivable rather than contract assets. Amounts are generally due within 30 days of invoice. At June 30, 2026, accounts receivable of $13,250,273 comprised $10,058,495 of billed and $3,191,778 of unbilled amounts; at September 30, 2025, the entire accounts receivable balance of $1,056,554 was unbilled and no billed amounts were outstanding, none of which was past due. No allowance for credit losses was recorded as of June 30, 2026 or September 30, 2025.

SCHEDULE OF ACCOUNTS RECEIVABLE, NET

As of June 30, 2026, total accounts receivable of $13,250,273 comprised $3,191,778 of unbilled amounts (approximately 24.1%), which have no invoice date and are presented separately, and $10,058,495 of billed amounts, of which approximately $3,673,089 (27.7%) was not yet due, approximately $4,020,675 (30.3%) was 1 to 30 days past due, approximately $2,362,828 (17.8%) was 31 to 60 days past due, and approximately $1,902 was aged more than 60 days. The foregoing percentages are expressed as a percentage of total accounts receivable, and the individual amounts may not sum to total billed accounts receivable due to rounding. The 31-to-60-day cohort of approximately $2,362,828 consisted almost entirely of three invoices to the Company’s two largest AI Inference Services customers that were invoiced early in the quarter as the billing cycle for the new offering was established; both invoices were collected in full in early August 2026, subsequent to period-end. The amounts aged more than 60 days consist of individually immaterial balances that management continues to monitor.

The Company evaluates expected credit losses on an individual receivable basis rather than on a pooled basis, because its receivables are concentrated among a limited number of customers with significant individual balances that do not share sufficiently similar risk characteristics to support pooling under the CECL model. In assessing expected credit losses, management considers each customer’s payment history and delinquency trends, the aging of outstanding balances, any known disputes or billing issues, current and forecasted economic conditions, and customer concentration. As of June 30, 2026, the Company’s two largest customers represented approximately 55.5% and 39.5% of total accounts receivable, respectively, or approximately 94.9% in the aggregate. The Company has not historically incurred material credit losses.

Although the Company’s accounts receivable is concentrated among a limited number of customers and the gross balance increased substantially during the quarter in connection with the ramp of the AI Inference Services offering, management concluded that no allowance for expected credit losses was required as of June 30, 2026 or September 30, 2025. Subsequent to June 30, 2026, the Company collected approximately $5.3 million of the accounts receivable balance outstanding as of June 30, 2026 (approximately $2.8 million in July 2026 and the remainder in early August 2026), including the entire 31-to-60-day past-due cohort described above, which cleared in early August 2026, corroborating that the past-due status reflected an administrative billing-and-collection timing lag associated with the launch of the new offering rather than credit deterioration. Management considered these subsequent collections as confirmatory evidence in its assessment and did not rely on them in establishing its measurement-date estimate.