I manage 10+ agents at Cisco. It doesn't mean longer hours, but the work can feel more intense.
Executive Take
Managing AI agents doesn't cut headcount needs for oversight, it shifts engineers from writers to reviewers. Leaders should plan for review bottlenecks and cognitive fatigue, not just productivity gains.
Executive Summary
Sergio Freitas, senior director of engineering at Cisco in Lisbon, manages 10 to 20 AI agents daily that write code and spawn sub-agents. He says the work isn't longer, but faster paced and more intense, since engineers now spend more time reviewing agent-generated code instead of writing it themselves.
Why It Matters
Technology and HR leaders need to know that agentic AI changes the nature of engineering work, not just its speed, creating new fatigue and skill gaps to manage. This affects how teams are structured and trained.
Bizquad Perspective
Most leaders count agent output as pure productivity gain, but the real bottleneck is shifting to human review capacity, which nobody is staffing for yet.