Reliance Global Group Completes Acquisition of Oblivus, Marking Its Entry Into the AI Infrastructure Market

Reliance Global Group completed its all-cash acquisition of Oblivus’s assets on November 3, 2026, entering the AI cloud and infrastructure market. Oblivus provides on-demand GPU computing for AI training and inference and operates a GPU marketplace, with approximately $1.5 million in annual revenue; Reliance says it remains debt-free and is discussing additional growth opportunities.

Analysis

The key underwriting question is whether this is a scalable, asset-light marketplace or a capital-intensive GPU hosting business. The distinction drives the outcome: marketplace volume can grow without matching balance-sheet expansion, while owned or committed GPU capacity can absorb cash quickly and expose returns to hardware obsolescence, utilization, and falling rental prices. “Debt-free” at closing is therefore not evidence that future growth is self-funded; the cash consideration also reduces liquidity, but its materiality cannot be assessed without the purchase price and cash balance.

At the disclosed revenue scale, the acquired operation is unlikely by itself to change consolidated fundamentals near term. Any valuation rerating would depend on verified customer demand and unit economics, not AI exposure. Larger cloud platforms and specialist GPU providers constrain pricing; a marketplace could differentiate through supply aggregation, but that advantage remains unproven. The strategic risk is execution across a new, technically demanding business, with potential distraction from the existing portfolio.

Near term, the completion removes closing uncertainty but leaves the more important diligence gaps open. Over 1–3 months, watch filings and management disclosures for acquired-asset scope, customer concentration, utilization, gross margin, capex commitments, and funding needs. Over 6–18 months, GPU price declines could improve customer economics while simultaneously compressing rental yields and shortening the useful life of owned capacity. The contrarian read is that the “AI infrastructure” label may attract attention faster than earnings can support it; absent operating evidence, neither rapid scaling nor durable marketplace economics should be capitalized into the stock.

Key Decisions for Investors

• Do not chase an AI-themed rerating on the closing announcement alone. Treat EZRA as a watchlist/event-driven name until purchase consideration, cash remaining, acquired assets, and post-close funding obligations are disclosed.
• Set a diligence trigger for the next filing or earnings update: require evidence of revenue conversion, utilization, gross margin, customer concentration, and capex per unit of capacity before considering a long. Falsification: growth accompanied by persistently weak margins or material new financing needs.
• For an existing position, size exposure against the possibility that GPU expansion consumes cash faster than revenue scales; reassess if management announces substantial equipment commitments, external financing, or a material shift in guidance.
• No clean pair or short is justified from the available facts. Revisit only if the stock materially reprices on AI positioning without corresponding operating disclosures, while checking liquidity and borrow availability before any tactical short.