Descriptions:
Ahnaf Priyo, senior engineering manager at Best Buy, presents the architecture behind the company’s agentic commerce stack at the AI Engineer conference, covering how AI shopping assistants are reshaping retail and what merchants need to build to participate. He cites data showing approximately 45% of all agent sessions on platforms like ChatGPT and Google Gemini are already shopping-related, and projects the agentic commerce market growing from $7 billion today to $65 billion by 2030.
The talk centers on two competing protocol standards: OpenAI’s Agent Commerce Protocol (ACP) and Google’s Universal Commerce Protocol (UCP), which define how AI agents communicate with merchants around product feeds, cart operations, and payments. Priyo contrasts their approaches to product data ingestion — both currently require merchants to push product feeds rather than supporting real-time catalog search calls, for indexing efficiency — and explains current payment limitations: ChatGPT uses shared payment tokens while Gemini UCP restricts payments to Google Pay. He highlights AP2, an extension of UCP, as the emerging standard for agent-authorized autonomous purchasing with spend caps and revocation controls.
The session includes a live demo using a Cerebras-hosted model running at 3,000 tokens per second, walking through a complete agentic shopping flow. Priyo also notes rapid ecosystem expansion, with Meta Commerce, Microsoft Copilot (UK Ray-Ban purchases), and a GoPuff/Grok collaboration all entering the space alongside the ChatGPT and Gemini incumbents.
📺 Source: AI Engineer · Published August 27, 2026
🏷️ Format: Keynote Launch
