Cisco Director Adapts to Managing AI Agents in Rapidly Changing Job

Sergio Freitas, senior director of engineering at Cisco in Lisbon, Portugal, shares how his job has changed significantly due to advancements in AI. He notes that software engineering has always undergone changes, but the current shift is rapid and caused stress among engineers.

The biggest change occurred after the launch of frontier models like Anthropic's Claude Opus 4.5 at the end of last year. This marked the beginning of embedding AI in workflows, allowing for delegation and collaboration with agents to do software engineering tasks. Freitas' role has shifted from manual coding to focusing on architecture, design, workflow orchestration, and managing these agents.

Engineers are now becoming system architects and workflow orchestrators, enabling them to solve more complex problems. Freitas manages 10-20 agents daily, which can spawn sub-agents for specific tasks. These agents continuously run in the background, aware of collaboration platform notifications, GitHub pull requests, and other integrations with explicit permission.

Freitas' team moved into loop engineering four months ago, where agents don't just ask questions but loop until they receive a final response. This has brought more work to engineers, as they need to review the code generated by AI. Although there's value in having AI collaborate with humans, strict guardrails remain to ensure accountability.