AI時代のエンジニア採用は「Why・What・Who」を考えられる人材へ | Sales Marker CTO 陳 晨
Sales Marker CTO Chen Chen explains how he built a 70-engineer team from 26 countries and why AI-era hiring should filter for Why/What/Who thinkers, not coders.
- Sales Marker’s engineering org grew from 4 founders to 70 engineers across 26 countries; language requirement was dropped to unlock a hiring positioning gap no other Japanese startup had claimed.
- Junior engineers are barred from using coding agents (Cursor, Claude Code, Windsurf) until they can code independently; all AI tool use requires story-point ROI measurement reported after the first month.
- Hiring criteria has shifted from hard coding skill to soft skills: can the engineer reason about Why we build this, What to build, and Who it is for — not just How.
- A new internal role, AI QA Engineer, generates both test cases and evaluations entirely via AI, replacing manual scenario writing that cannot scale for AI-native products with unbounded input space.
- Orcha, Sales Marker’s AI orchestrator product, targets the Japanese enterprise market specifically because major LLM players have not penetrated it due to quality and security requirements.
- Data is the one moat technology cannot commoditize: Chen tracks proprietary data asset creation and alliance-building as among his most important CTO responsibilities.
- Chen defines the CTO role as a business operator first — he directly owns product, hiring, and back-office at Sales Marker — and the ideal CTO continuously finds company bottlenecks and solves them with technology.
2025-12-01 · Watch on YouTube