You will build, deploy, and maintain frontier AI agents for our enterprise customers using the most advanced browser and computer-use technology in the industry.

Asteroid is the first company to deploy computer-use agents in healthcare at scale. Our agents execute thousands of mission-critical tasks every day, automating workflows that no one else can touch: complex multi-step portals, legacy systems with no APIs, exception-heavy processes that require real-time judgment.

As we onboard more enterprise customers, each with unique workflows and edge cases, we need someone embedded in the delivery process who can take a customer's workflow from scoping to production. You'll be the bridge between our customers and Asteroid platform, translating messy, real-world processes into reliable, high-performing agents.

Why This Role Exists

As our customer base grows, we need a dedicated forward deployment engineer who can own the full agent lifecycle for assigned accounts, from initial scoping through production monitoring. You'll work directly with customers in shared Slack channels, understand their workflows deeply, and build agents that handle the complexity and edge cases of real enterprise operations.

You'll be working with frontier computer-use and browser-use technology every day. You'll have a direct line to the product and engineering team, and your feedback will shape the direction of a fast-growing platform that's defining a new category.

What You'll Own

• Account Management. Be the dedicated point of contact for your assigned customers. Own the relationship day-to-day in shared Slack channels. Keep them informed, unblocked, and successful.
• Agent Scoping & Requirements. Join customer calls. Understand workflows end-to-end: what the human does today, where the edge cases are, and what "done" looks like.
• Agent Building. Translate customer workflows into prompts, tool configurations, and agent logic on the Asteroid platform.
• Testing & Iteration. Run agents against real scenarios. Catch edge cases. Iterate with the customer until the agent reliably does what it should.
• Production Monitoring. Monitor deployed agents. Investigate failures. Fix prompt, models or configuration issues.

What You'll Contribute To

• Product Direction. You'll be closer to the agent-building experience than anyone. Surface friction and suggest features. Your feedback goes directly to our product and engineering team.

This Role Is for You If

• You're excited about working with frontier AI technology that no one else has deployed at this scale.
• You can take a vague customer requirement and turn it into a precise, working agent.
• You're comfortable working directly with customers, explaining what's possible, setting expectations, and iterating together.
• You're a fast learner who can pick up new tools and platforms quickly.
• You think in systems. When an agent fails, you diagnose whether it's a prompt issue, a tool issue, or a workflow design issue.
• You're self-directed. You manage your own time, communicate proactively, and ship.

Qualifications

• Strong understanding of LLMs, prompt engineering, and how to get reliable outputs from AI systems.
• Able to read and understand LLM-generated JavaScript and Playwright code. Our agents write and execute code autonomously, and you'll need to review, debug, and guide that output.
• Familiar with APIs, webhooks, and how web applications work under the hood.
• Excellent spoken and written communication. You'll live in Slack, take customer calls, and need to be clear and responsive.
• Able to work independently across time zones.

Nice to Have

• Experience building computer-use or browser-use AI agents.
• Understanding of healthcare processes and workflows (prior auth, insurance, EHR systems, claims).
• Experience in a customer-facing technical role (solutions engineering, technical consulting, forward deployment).

Who Should Apply

We're open to all experience levels. University students, new grads, and career switchers are welcome. What matters is that you're sharp, technical, and excited about this work. If you've been building side projects with LLMs, automating workflows for fun, or tinkering with agents, we want to hear from you.