On December 8, 2024, The Washington Post examined the debate around “ghost engineers”—software engineers whose measured contribution appeared unusually low—and the challenge of assessing developer productivity responsibly.
The story identifies Simon Obstbaum as an entrepreneur and former chief technology officer of Crunchyroll. It describes the machine-learning approach as conceived in collaboration with Stanford researcher Yegor Denisov-Blanch and associate professor Michal Kosinski. Rather than relying only on raw activity counts, the method evaluates the effect of individual and team work on a company’s broader codebase.
The article also highlights the limits of simplistic measurement. Engineers may contribute through architecture, mentoring, research, or other work that produces little code, and low measured output can reflect unclear roles or broken incentives. The broader research goal is to understand these patterns and help organizations improve them—not to reduce engineering work to a single activity metric.
