Nixon Dinh on Product Catalogs for AI Agents

AI Engineer16m 5s
Nixon Dinh beside the headline AI Shopping Search.
0 comments · 0 votesOpen discussionClose discussion
Sign in to join the discussion

    PayPal product director Nixon Dinh discusses how shopping agents might find products using intentAgentic commerce uses AI agents to assist with stages of buying and selling, such as product discovery, comparison and checkout. and how this differs from keyword search. He contrasts literal matching with semantic retrieval and describes a hybrid approachHybrid search combines retrieval methods, commonly keyword and meaning-based search, so AI applications can handle both exact matches and paraphrased requests..

    Dinh presents PayPal’s catalog enrichment experiments, reporting that added product detail improved recommendations in its tests while excessive irrelevant text sometimes increased hallucinationsA hallucination is an AI output that presents false, unsupported, or invented information as though it were reliable.. The recording does not provide a quantified evaluation protocolEvaluation is the systematic assessment of a system's behaviour or results against explicit criteria., so the results are attributed to PayPal.

    Original YouTube thumbnailWatch on YouTube

    Video topic coverage

    • Agents & workflows85
    • Coding & tools0
    • Models & learning35
    • Infrastructure & compute35
    • Research & evaluation65
    • Safety & security0
    • Business & productivity100
    • Creative & media0
    • Robotics & perception0
    • Society & policy0

    Share this video

    Nixon Dinh beside the headline AI Shopping Search. With WWW.ARTIFICIAL-INTELLIGENCE.VIDEO, 4 October 2026 and duration 16m 5s.

    Nixon Dinh discusses how product catalog data can support AI shopping agents.