Back to careers

Agent Engineering

Agent Engineer

Fixed Labs is building an agent platform for businesses that want AI to perform real work alongside their teams. As an Agent Engineer, you will embed inside the organizations we work with, learn how their operations run, and deploy and manage agents that handle customer service and other recurring business workflows. Your job is not finished when an agent goes live: you will own its reliability, results, and continuous improvement in production.

Who we are looking for

Above everything else, we are looking for high achievers: people who are exceptionally resourceful, learn unusually fast, and have a history of making difficult things happen.

You do not need the perfect résumé or experience at a famous technology company. We are not optimizing for years of experience. We care about evidence that you are exceptionally good at what you do.

  • Built and shipped ambitious products while you were still in school.
  • Became one of the strongest performers on a previous team.
  • Started a company or a serious side project.
  • Earned technical, academic, entrepreneurial, or competitive achievements.
  • Contributed meaningfully to open-source projects.
  • Taught yourself difficult technical skills unusually quickly.
  • Automated something everyone else accepted as manual.
  • Took ownership of a problem well outside your job description and solved it.
  • Built something people actually use, or developed another track record that demonstrates exceptional ability.

The role

As an Agent Engineer, you will own AI agents operating inside customer organizations. You will embed with client teams, understand how their work actually happens, and turn recurring processes into reliable agent workflows.

You will build agent skills, connect tools, create workflows, test behavior, deploy agents, monitor what they do, investigate failures, and continuously improve them. If an agent fails, you own understanding why. If it meets a new edge case, you determine how it should respond. If it needs another system, you work out the integration.

Your responsibility is not simply to ship an agent. It is to make the agents you own successful at their jobs, including customer service and other operational functions.

What you will do

  • Build, configure, deploy, and manage AI agents on the Fixed Labs agent platform.
  • Create agent skills, tools, instructions, context, guardrails, escalation paths, and workflows.
  • Connect agents to APIs, SaaS applications, databases, MCP servers, communications platforms, and customer systems.
  • Build integrations and automations when existing capabilities are not enough.
  • Deploy and operate agents for customer service and other recurring operational work.
  • Test agent behavior and create evaluations for accuracy, task completion, autonomy, and escalation quality.
  • Monitor production agents and investigate failures across trajectories, tool calls, integrations, context, and workflows.
  • Identify recurring failure patterns and build permanent solutions.
  • Continuously improve agent reliability and autonomy.
  • Work directly with customers to understand how their businesses operate.
  • Translate customer problems into technical solutions without waiting for someone else to define every requirement.
  • Identify additional opportunities for agents to take on useful work.
  • Turn lessons from individual deployments into reusable platform capabilities and operating playbooks.

What we value

  • High agency — You do not wait for perfectly written requirements. You investigate problems, learn what you do not know, experiment when documentation is weak, and find another route when the obvious solution fails. You know when to ask for help, but ambiguity makes you start figuring things out.
  • Problem solving — Agent engineering regularly involves strange behavior, undocumented APIs, customer-specific systems, ambiguous business rules, and edge cases nobody anticipated. We want people who enjoy that kind of work.
  • Learning speed — The AI stack changes quickly. What you already know matters; how quickly you can learn what you do not know matters more.
  • Ownership — If something you deployed is not working, “that is not my part of the system” is not an acceptable answer. You work across boundaries, understand the problem, and involve the right people to solve it.

Technical foundation

We do not expect you to know everything. We expect you to be capable of learning extremely quickly and to be comfortable working across unfamiliar systems.

  • Python and/or TypeScript or JavaScript.
  • APIs, JSON, webhooks, authentication, and integrations.
  • Git and basic software engineering practices.
  • LLMs, modern AI systems, and AI coding tools.
  • Debugging unfamiliar systems and reading technical documentation.
  • Experience with agents, MCP, browser or computer use, workflow automation, databases, cloud infrastructure, or LLM evaluations is a plus.

What success looks like

Your agents become increasingly autonomous and reliable. Customers trust the agents you own. You solve problems without needing someone to prescribe every step, and you identify opportunities other people miss.

Six months after joining, you should be capable of solving problems we would not have trusted you with on day one. We want people with a steep trajectory. If you are early in your career but have repeatedly demonstrated that you learn faster, build faster, and take more ownership than the people around you, we want to hear from you.

Job conditions

  • Remote.
  • Compensation: paid in USD.
  • English level: advanced.
  • Full-time, Contractor.
  • Schedule: 9:00 AM to 5:00 PM Eastern Time, Monday through Friday.
  • Paid time off: two weeks per year.
  • The role begins with a two-month paid evaluation period under a full-time contractor agreement.
  • At the end of the two-month evaluation period, continued long-term, full-time employment will be considered based on individual performance and current business needs. Completing the evaluation period does not guarantee continued employment.

Apply

Send your application

All fields are required so we can review your application without chasing missing context.