PulseAugur
EN
LIVE 03:06:05

JD.com launches industrial LLM/VLM platform for e-commerce item management

JD.com has developed the JD Oxygen AI Item Center (Oxygen AIIC) V1, an industrial-scale platform designed to manage and understand items for e-commerce using large language and vision-language models. The system addresses challenges in knowledge production for billions of products by employing ontology engineering, a "Semantic Search then Discrimination" architecture, and self-evolving AI models. Oxygen AIIC achieves 94.2% precision and 82.8% recall, processes hundreds of millions of item updates daily, and has demonstrated significant improvements in search traffic coverage and item information quality. AI

IMPACT This platform demonstrates how large-scale LLMs and VLMs can be applied to solve complex data management challenges in e-commerce, potentially influencing similar systems in other retail sectors.

RANK_REASON This is a product launch from a major e-commerce company detailing an internal platform, not a frontier AI lab release.

Read on Hugging Face Daily Papers →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

JD.com launches industrial LLM/VLM platform for e-commerce item management

COVERAGE [1]

  1. Hugging Face Daily Papers TIER_1 English(EN) ·

    JD Oxygen AI Item Center (Oxygen AIIC) V1: An Industrial-Scale LLM/VLM-Centric Solution for Item Understanding, Management, and Applications

    A large-scale industrial platform leveraging LLMs and VLMs addresses key challenges in structured item knowledge production for e-commerce, achieving high precision and throughput across billions of products.