Moshi Wei

AI PhD building practical agent systems, developer tooling, and applied LLM products.

Highlights

• Delivered high-impact LLM products at LRQA (a Goldman Sachs company) and Huawei
• Founder of ISEK, an open-source multi-agent framework with 560+ GitHub stars and a 600+ member community
• Core contributor to ANP, an agent network protocol focused on interoperable agent systems
• Builder of AutoAide, a persistent assistant runtime for Codex
• Guest speaker in AI communities including MLLM, Datawhale, and WaytoAGI

Selected Work

• AutoAide Open-source persistent assistant runtime that gives Codex a real runtime home: dedicated workspaces, session continuity, shell execution, and channel integrations.
• ISEK Open-source decentralized multi-agent framework for orchestration, tool execution, state management, and collaborative agent systems.
• ANP Agent network protocol and SDK work for connectivity, coordination, and scalable execution across interoperable agent environments.

Industry Impact

• Built LLM scheduling optimization systems with projected global workforce impact at LRQA
• Shipped an audit report generator that reduced report writing time by 83% for 700+ users
• Built automated audit review tooling that replaced 30+ manual roles in production workflows

What I Work On

• AI agents and agent runtimes
• Applied LLM systems and workflow automation
• Developer tooling and agent infrastructure
• Multi-agent protocols, orchestration, and execution systems

Looking For

High-leverage remote roles in AI products, agent infrastructure, developer platforms, or applied LLM systems.

Connect

• GitHub: @Moshiii
• LinkedIn: https://www.linkedin.com/in/moshi-wei/
• Google Scholar: https://scholar.google.ca/citations?user=MoshiWei
• Email: wmswms938@gmail.com
• Website: https://moshiii.github.io/