Microsoft and CMU Launch PRAISE to Rebuild What AI Took From Junior Engineers
When AI can write the code, what does a junior engineer actually do?
Microsoft and Carnegie Mellon University published the answer they’re betting on October 15: the Preceptorship for AI-accelerated Software Engineering program, or PRAISE. The initiative pairs selected early-career Microsoft engineers with senior engineering preceptors and CMU coursework, borrowing the structure from nursing, where new nurses shadow experienced practitioners before working independently.
The program was conceived after Mark Russinovich, Azure’s CTO and a CMU alumnus, visited the university’s Kenneth C. Griffin School of Computer Science in October 2025. Russinovich’s observation was specific: AI tools are compounding the leverage of experienced engineers, but that same leverage isn’t available to people who haven’t yet built the underlying judgment.
“AI changes the work, but it doesn’t eliminate the need for junior engineers or for human judgment and accountability,” Russinovich wrote in the program announcement.
The Structural Problem
Senior engineers know what to build and when to use AI to build it. Junior engineers are supposed to develop that taste by doing the less-complex work that senior engineers no longer bother with. AI has absorbed much of that work.
The concern isn’t that junior engineers are becoming unnecessary. It’s that the traditional on-ramp — doing the task work, making the mistakes, learning from the senior engineer in the next seat — is disappearing before a replacement has been designed.
Michael Hilton, teaching professor and associate department head at CMU’s Software and Societal Systems Department, put it plainly in a statement accompanying the announcement: “It used to be that early-career software engineers would do a lot of tasks that would build up their expertise.”
PRAISE is an attempt to recreate those developmental conditions deliberately, rather than hoping they emerge from whatever work AI leaves behind.
What the Program Does
Preceptees are selected early-career Microsoft engineers. They are paired with a senior preceptor — an experienced engineer who coaches them through real project work, shares their approach and decision-making process, and flags when to use AI and when not to. The arrangement runs alongside CMU coursework.
This semester’s curriculum includes Machine Learning in Production and Preceptorship in AI-Augmented Software Engineering I. Spring courses are planned in DevOps and a second preceptorship course. The CMU credits can count toward a master’s degree in software engineering.
Scott Hanselman, VP at Microsoft, co-authored the program announcement. Jace Watson, a Seattle-based Microsoft software engineer who enrolled in the program, describes joining right as AI tools became embedded in day-to-day work — he wanted structured guidance, not just tool access.
What’s Missing
Neither Microsoft nor CMU has published participant numbers, selection criteria, or measured outcomes. The program was publicly described in Microsoft’s September 17 account of its own AI transformation, but outcome data — retention, productivity, engineering quality — has not been shared.
PRAISE is a structural bet: that pairing junior engineers with senior mentors and formal curriculum is a better substitute for the task-based learning AI is eroding than doing nothing. That may be right. It may also be a well-designed program that does not survive contact with the scale of the problem. One named participant and a course list are not a verdict.
CMU says it hopes to expand to more students and eventually other industry partners.