From code to context

Justin Johnsen joined KPMG six years ago and has been a forward-deployed engineer for about a year, working directly with clients from understanding business needs through delivering solutions.

What the job has become

Johnsen says much of his work now is AI. He gathers information for AI to generate valuable assets, spreads outputs across client deliverables, and coaches client teams on using AI effectively.

“AI is incredibly capable, but it's really only as good as its context.”

Speed through preparation

For one client, Johnsen was hired to integrate applications into a complex system. After months gathering requirements and refining context, the team used AI to deliver the application in about a month.

The new bottleneck

AI has let Johnsen multiply his output and maintain context across multiple clients simultaneously. The bottleneck used to be software development and writing code; now it is gathering context and refining intent before development begins.

Knowledge transfer in real time

A forward-deployed engineer offers more than a traditional software handoff. By working directly with client teams while developing, training, and coaching, Johnsen transfers knowledge in real time and delivers best practices alongside the application.

Human judgment takes priority

AI gives Johnsen more space to define problems and make critical decisions instead of spending time on repeatable, monotonous tasks. He can focus more on things that require human judgment.