← Browse

Europe 2026

Why (Senior) Engineers Struggle to Build AI Agents — Philipp Schmid, Google DeepMind

Philipp Schmid

Overview

Building AI agents presents unique challenges compared to traditional software development. Instead of acting as traffic controllers with predefined rules, engineers now function more like dispatchers, defining goals for agents without dictating every step. This shift requires embracing the non-deterministic nature of AI, treating errors as inputs, and moving from rigid unit tests to broader evaluations of agent reliability and success rates.

Who should watch

Key takeaways

Notable quotes

Philipp Schmid notes that when building agents, *we define the goal on what we want the agent to do, but we don't define the exact step the agent needs to take to achieve that goal.*
Schmid also states, *We need to move from unit test to evals. We need to test how often something works because agents are only successful if they are really reliable.*

Watch on YouTube →

Unofficial community note. Prefer the recording for nuance.