McKinsey explainer: three McKinsey leaders on what FDEs do, how they differ from consultants, and how AI is changing the job

McKinsey published a video Explainer on October 16 in which three of its leaders, Holger Hürtgen, Martin Harrysson and Nick Robson, set out what a forward-deployed engineer does. They describe an engineer doing data science, data engineering and software engineering directly on a client's systems rather than in a lab, with the goal of reaching an outcome, not just getting software to run. Hürtgen concedes the term may be a bit of a buzzword, but argues the concept matters more than ever because LLMs on their own don't produce bottom-line impact unless they are tailored to a specific problem.

On the consultant question, the three don't fully agree: Harrysson and Hürtgen say an FDE leaves behind a working product while a consultant leaves a recommendation, whereas Robson says the two roles are merging and that a really good consultant makes a really good FDE. Robson also says demand for QuantumBlack, McKinsey's AI arm, is higher than ever, with the longest wait list he has seen for that kind of talent.

They say AI has changed the job in two ways: engineers write less code and spend more time on specifications because agents do much of the coding, and the product itself is increasingly an agentic AI system that needs a model, a data layer, a harness and changed workflows. The first weeks of an engagement, they say, are mostly spent sitting with the people doing the work, scoping the problem and finding the data.

For job seekers, the piece is a useful read on how a large consultancy frames the role: technical enough to understand systems, but hired just as much for communication, quick learning and the ability to get people to actually change how they work.