Data/ML/AIEmergingAround since 2024
Computer Use / Browser Agents
Also written as Computer Use, Browser Automation Agent, Agentic Browsing
AI agents that operate a screen or browser directly — clicking, typing, reading pixels or the DOM — to complete multi-step tasks, instead of only calling predefined APIs.
Think of it like
Handing someone the mouse and keyboard and a goal, instead of giving them a fixed set of buttons that already do what they need.
Junior or senior?
Very new and still brittle in practice.
Senior sounds like
Talks about failure recovery and guardrails on what the agent is allowed to click, not just the demo working once.
Ask them
“What happened when the agent clicked the wrong thing or got stuck? How did you catch it?”