Podcast Atlas
concept Updated 2026-08-11 Tags: Enterprise-Ai, Fde, China, Deployment

Chinese-Style FDE / 中国式 FDE

Chinese-style FDE is the source’s name for [[ForwardDeployedEngineer|FDE]] practice under Chinese enterprise conditions, where deployment often includes rebuilding the data, workflow, permission, and organization substrate before AI agents can perform. In E248｜一个“催发货”AI要跑通260步，和阿里瓴羊彭新宇聊聊中国式FDE, [[PengXinyu|彭新宇]] argues that the role is scarce because it combines AI acuity, industry depth, and data breadth, and because it must deliver business effect rather than software functions.

The concept extends prior FDE realism from E240｜OpenAI联手PE砸下40亿美元，聊聊硅谷最火新职位FDE and EP128 从 Palantir 到 OpenAI：FDE 会成为 AI 时代最重要的新岗位？ 🧬. Those sources emphasized customer-site discovery, workflow triage, and product-feedback loops. The 瓴羊 episode adds that in China, weaker workflow and data standardization can make the job closer to participating from foundation work through production rollout.

Key Claims

  • Chinese-style FDE starts from business outcomes such as service efficiency, marketing ROI, sales productivity, and conversion, not a requested feature list.
  • It often has to repair China Enterprise AI System Debt by clarifying business data, process steps, support libraries, permissions, and exception handling.
  • The role is not one person by default; [[Lingyang|瓴羊]] describes a team made of business analysts, AI architects, and customer-side expert coaches.
  • Good deployment candidates are high-volume workflows where cost, time, or labor pain is visible and acceptance can be measured.
  • Top-down sponsorship matters because the budget case often requires adding IT spending and business spending into one total ledger.
  • Model choice is secondary to operational reliability: system deployment, model deployment, gray rollout, evaluation, and human confirmation all need separate controls.

Connections