At a senior engineering conference, discussions centered on AI's impact on teams, junior engineer development, and managing agentic systems. A key challenge emerged around controlling agent access to third-party tools like Notion: building gateways with distinct agentic identities and token exchanges to enforce granular permissions that the third-party provider itself may not support.
An engineer reflects on learning that influence requires more than technical correctness—it demands understanding people's beliefs, values, and context, then advocating skillfully. The author shares insights from speaking at LDX3 and invites readers to a workshop on building influence beyond technical expertise.