This article explains how Alibaba engineers transformed millions of HSF microservices into MCP servers without code changes by leveraging the Higress AI gateway, detailing the ecosystem components, step‑by‑step migration process, challenges of rapid MCP evolution, and strategic insights for AI infrastructure teams.
At the 2024 Yunqi Conference, Alibaba Cloud’s AI Infra experts detailed the latest challenges of large‑model deployment—such as hardware costs, resource management, and software‑hardware coordination—and introduced PAI’s new capabilities, including stability tools, automated distributed training, reinforcement‑learning frameworks, inference optimizations, and integrated big‑data AI solutions.
This article describes the Hybrid Embedding solution proposed by Ant AI Infra to address storage, resource, and feature‑governance challenges of massive sparse CTR models, detailing its multi‑layer storage design, KV‑based parameter server, and performance gains in large‑scale recommendation systems.
