EmergingTool-using agents and workflow orchestrationPart of The agent workforce

Computer-Use Agent Engineer

Also called Computer Use Engineer · Computer-Use / Browser Agent Engineer · GUI Agent Engineer

A Computer-Use Agent Engineer builds agents that operate real application interfaces, clicking, typing and reading screens, within firm limits. They make automation possible where no clean API exists, without putting sensitive actions at risk.

What they do

They build agents that read an interface, decide the next step and act on it through explicit tool permissions. Much of the day goes on interpreting UI state: what is on screen, what changed, whether the last action worked. They write verification steps after actions, set boundaries around sensitive operations, and run agents in sandboxes. They assemble a representative task suite, run it often, and study failures. They connect agents to tools through agent frameworks and protocols such as MCP, and work with workflow engines when tasks need several stages. Much of the work is debugging odd interface behaviour.

What a strong hire delivers

In the first months, a strong hire ships a working agent that completes a defined set of real tasks in a sandbox, with clear rules on which actions need approval or are blocked. They also leave behind a task suite the team trusts, a failure log sorted by cause, and a repeatable way to check whether each change helped or hurt.

How you can tell they are good

Ask them to walk through an agent they built and where it failed. Strong people describe how the agent confirms an action worked, not just that it was sent. They can explain their permission model, show a task suite, and say what they chose not to automate. Be wary of demos that only show the happy path.

Why it is hard to fill

The title barely exists yet. The work sits inside roles called automation engineer, test engineer or applied AI engineer, so searching by title finds few people. The mix is unusual: interface automation, vision models, evaluation and security thinking. Few candidates have all four, and those who do are usually busy. Scarcity is high.

Where to find them

Look first at former RPA and test-automation engineers, who already know how fragile interfaces are. Application engineers and systems integrators who moved into browser automation are another pool. Others come from eval and tooling teams. Typical employers are AI labs and automation startups, though some work happens quietly inside product teams at larger software firms.

Indicative pay

Not enough published evidence yet to give a fair range for this role, which is common for work this new. We benchmark it for you as part of a search.

Interview questions for a Computer-Use Agent Engineer

  1. Describe a computer-use agent you built. How did it know each action had actually worked?

    Look for concrete verification: checking screen state, expected elements or side effects after each step, with retries and escalation. Worry if they rely on the model saying it succeeded, or have never measured false successes.

  2. How did you decide which actions the agent could never take without human approval?

    Strong answers describe classifying actions by risk, explicit permissions, allow lists and approval gates enforced outside the model. Worry about boundaries held only in the prompt, or no clear list of sensitive actions.

4 more questions, with what a strong answer covers, are in the Role Kit.

Job description you can post

We are building agents that work through real application interfaces, within strict limits on what they may do. We need an engineer to design, test and harden them. You will own the agent and the task suite that tells us whether it can be trusted.

What you will do

  • Build agents that interpret application interfaces and take actions through explicit tool permissions
  • Design checks that verify each action had the intended effect
  • Define and enforce boundaries around sensitive actions
  • Create and maintain a representative task suite and run it regularly
  • Analyse failures and turn them into fixes and new test cases
  • Integrate agents with tools, workflow engines and protocols such as MCP
  • Run agents safely in sandboxed environments

What you will bring

  • Built and shipped browser, desktop or interface automation that others relied on
  • Experience reading and handling unreliable UI state, such as timing and layout changes
  • Practical experience evaluating model or agent behaviour with a test harness
  • Clear thinking about permissions, security and failure modes
  • Working knowledge of agent frameworks and tool integration
  • Able to explain trade-offs and past failures plainly

Nice to have

  • Background in RPA or test automation
  • Experience with vision models for screen understanding
  • Experience building sandboxes or isolated test environments

Free to use and adapt. If it saves you time, a link back to this page is appreciated.

Skills that matter

  • Agent orchestrationAdvancedCoordinate tool-using agent execution
  • Tool and protocol integrationAdvancedConnect models to external data and capabilities
  • Security engineeringWorkingProtect systems, identities and sensitive assets
  • Model evaluationWorkingMeasure capability and failure modes credibly

Other roles in tool-using agents and workflow orchestration

Hiring nowAI Agent EngineerAn AI Agent Engineer builds software agents that take real actions through tools they are explicitly permitted to use. They turn model capability…Hiring nowAgent Tool Integration EngineerAn Agent Tool Integration Engineer connects AI agents to outside systems through tightly scoped interfaces. They make sure agents act only where…EmergingAgent Workflow ArchitectAn Agent Workflow Architect decides which tasks go to agents, which stay with people, and where approvals sit. It matters because agents that act…EmergingAgent Memory and Personalisation EngineerBuilds the memory, user models and personalisation layers that let an agent remember people across sessions and act on it. It matters because an…EmergingAI Agent ManagerAn AI Agent Manager assigns work to AI agents, reviews what they produce, handles the exceptions and tunes the workflow. They make agents dependable…

Related roles in other families

EmergingMulti-Agent Systems ResearcherMulti-agent coordination and economicsA Multi-Agent Systems Researcher studies how several AI agents behave when they work together, and builds methods and benchmarks to make that…EmergingAgent Coordination EngineerMulti-agent coordination and economicsAn Agent Coordination Engineer builds the layer that lets several AI agents hand work to each other, share state and settle conflicts. It matters…EmergingMulti-Agent Evaluation ScientistMulti-agent coordination and economicsA Multi-Agent Evaluation Scientist measures how groups of AI agents behave together and designs tests that expose collusion, cascades and…EmergingAgent Market Mechanism ResearcherMulti-agent coordination and economicsThis person designs and tests the rules that govern how AI agents trade, bid and share resources in bounded simulated markets. It matters because…

How we wrote this

Role definitions are our own editorial work, checked against occupational classifications, standards, research and employer pages. Titles vary by company, and this is a description of the work, not a job advert.

← Tool-using agents and workflow orchestrationAll new job families

Start a search

Hiring a Computer-Use Agent Engineer?

Tell us the work. We find the people already doing it, whatever their job title says, and you meet them already interviewed for your role.