Agent Advocacy: Why Developer Relations Must Rebuild for AI Users

The traditional developer advocate role, built on a two-way feedback loop with human developers, is becoming obsolete. Stephanie Jarmak argues that AI agents are now both users of developer tools and recommenders of them, fundamentally changing the DevRel function.

This shift creates a new role: the agent advocate. This person measures how agents interact with tools through benchmarks like CodeScaleBench, analyzes failure traces, and optimizes agent friction. They also manage generative engine optimization (GEO) to ensure products are discoverable by agents.

The change is organizational, not just technical. Teams must decide where agent advocacy sits – engineering, product, or marketing – because it spans all three. The core DevRel mission remains, but now includes educating both humans orchestrating agents and the agents themselves.

Tradeoffs are clear: agent advocacy requires new instrumentation and evaluation infrastructure, but it yields tight feedback loops impossible with human users. The risk is over-investing in agent-facing content while neglecting human developer experience, or treating agent recommendations as a passive SEO problem rather than an active design challenge.