Software Engineer, Agent Platform (all levels)

About Stepwork

We’re Stepwork. We’re rethinking how companies automate repetitive work, starting with IT teams. Instead of building endless APIs and integrations, we teach software to use other software the way a human would, clicking through UIs and making decisions in real time.

Today, companies pay entire IT teams to manually provision software licenses, onboard and offboard employees, and run access reviews for audits. As software sprawl grows, this work only increases.

We’ve built automations that already handle these workflows. We have pilot customers like Calm, Snowflake, and Ramp, with more than 35 large enterprises asking to get started. We’re small, the demand is real, and what we build next will set the bar for this category.

What You’ll Work On

This role is fullstack with a strong backend + infrastructure emphasis. You’ll own real systems end-to-end, shipping quickly while maintaining reliability expected by enterprise customers.

You’ll be building the execution environment for automation. Our Electron-based application runs a secure virtual machine on the user’s device, executing packaged “flows” with extreme accuracy. You’ll extend and harden this system, improving speed, reliability, and testability.

• flow orchestration
• prompt packaging + execution
• VM lifecycle management
• reliability testing across real + virtual machines
• scaling the system as we transition from just local to local and customer's own cloud

🧪 Testing & Reliability Systems

Today, our bottleneck is how quickly and accurately we can test flows. We target 98% reliability (industry average: ~60%). You’ll work on systems that test flows across multiple environments—real hardware, virtual machines, and cloud-hosted compute.

💥 Customer Impact

We work directly with customers like Calm, AngelList, Affirm, and others. Their environments are complex and often undocumented. When a workflow breaks, they expect it to work—regardless of quirks in their browser, VM, or internal tools.

Your work ensures it does.

This Role Is Essential Because…

We have more enterprise demand than we can service. Adding two engineers unlocks our ability to onboard ~20 additional logos and meaningfully expand ARR.

How You’ll Work

This role is built for engineers who:

• Move quickly without breaking the fundamentals
• Prefer practical solutions over architectural ceremony
• Enjoy breaking down vague problem spaces
• Take ownership from “I’ll figure it out” → “it’s shipped and reliable.”
• Communicate clearly and ask strong questions
• Thrive in a small, in-person team where decisions happen face-to-face

What You’ll Bring

Technical Experience

• 4–10 years building production systems (fullstack or backend)
• Strong backend engineering foundation (Go + TypeScript)
• Infra experience with Docker, AWS, Terraform
• Comfort with VM debugging, local-device constraints, or adjacent systems
• Ability to pick up light frontend work when needed

Capabilities We Care About

• Ownership: If you ship it, you own it, good and bad.
• Principled Problem Solving: You ask the right questions, avoid scope creep, and choose the simplest path that works.
• Persistence: Many problems you’ll encounter don’t have documentation. You enjoy figuring them out.
• Speed: You know how to ship quickly without overengineering.
• Disagree & Commit: You can push back, align, and move.

Nice-to-Haves

• Experience working on Electron or similar desktop runtimes
• Infra-heavy product experience
• Previous work on early engineering teams

Compensation Range: $125K - $200K