Surprising lessons from my research scientist job search
A research scientist shares unexpected lessons from their job search, including that networking and soft skills matter as much as technical expertise, company culture and research alignment are critical factors, and the process often reveals mismatches between academic training and industry expectations.
Background
- The author, Yongzhe Zhang (alias "yongzx"), is a PhD graduate in machine learning / systems who published this blog post in June 2026 documenting his experience searching for research scientist roles.
- Research scientist (RS) is a distinct job category in tech (especially AI/ML) that sits between academic faculty and industry software engineer—focused on publishing, experimentation, and advancing methods rather than shipping products.
- The post surfaces structural realities of the 2025–2026 AI job market: supply of PhD grads has surged, many top research labs have slowed hiring or shifted toward applied roles, and the "research scientist" title now means different things at different companies.
- Key dynamics: internal transfers from engineering roles into research teams are common but hard; conference networking and publication track record often matter more than leetcode-style interviews for pure RS; many companies have replaced open-ended research interviews with take-home tasks or system-design rounds.