The most valuable thing in the future won't be parameter-tuning experts, but rather versatile individuals who can embed Claude Code into the company's core operations and understand business pain points.

This article focuses on AI Frontline Deployment Engineers (FDE) and their value, emphasizing that deploying general large models as enterprise-customized agent workflows requires on-site execution and communication. The author predicts that in the future, the greater demand will be for AI engineers within companies who are proficient in prompts, frameworks, and evaluation, and skilled in using tools like Claude Code, Codex, and others to embed AI capabilities into software and business. AI roles will diversify into areas such as LLMOps, evaluation engineers, and AI data engineers, with truly scarce talent being versatile individuals who understand both engineering and business.