Software Engineer, Workload Orchestration (Distributed Compute)
Posted Updated
Description
About nTop
nTop builds computational design software for engineers designing next-generation aircraft, defense systems, and performance vehicles. Our platform uses fundamentally different approaches to geometry representation and computation, enabling designs traditional CAD systems cannot express. We're moving from selling seats to selling scaled compute, and our Orchestrate team is building the distributed computing platform that makes that possible.
The Role
You'll be a senior technical leader on Orchestrate, working as a peer to our Orchestrate tech lead. This is a senior technical leadership role, not a people-management role: you won't have direct reports. You'll lead a team of AI agents, with the impact of a much larger team. You decide what gets built and how, set the technical direction and standards, and drive the work to production through agents instead of through people.
We want someone who knows how to run a software project and is excited to run it this way. You'll own open-ended distributed-computing problems end to end, from researching the options to shipping the result, and you'll make sure the work the agents produce is the right work and is good enough to ship.
Orchestrate is one of the most important areas of investment at nTop over the next year. Our product is moving from an installed Windows application to a collaborative, multiplayer, agent-driven engineering workflow that runs mostly on the server, and compute is a major part of it. Orchestrate is also where we are working out how agent-led development should work, so you'll help shape that practice for the rest of the company.
What You'll Own
- Open platform problems (1-3 quarters). Questions the team knows matter but can't yet answer, such as how to run workloads reliably across customers' on-prem, private cloud, and government cloud compute, or how to move from long-running monolithic tasks to a scheduler that survives failures. You research, prototype, compare trade-offs, and recommend a path, including "not yet."
- Which problems come first. You help decide which problems to solve, in what order, and what to leave alone. You are accountable for whether those were the right bets, not only for how well they were executed.
- Work done through agents. You break projects into tasks agents can carry out, review what they produce, and keep quality high. You own the result as if you had written it yourself.
- Architecture and standards. You set the technical direction and engineering standards for the part of Orchestrate you lead, and you keep them coherent with the rest of the platform.
- Two agent teams, one goal. You and the other Orchestrate team lead each lead an agent team toward shared, ambitious goals for the year. You design together how the teams divide work and collaborate.
- Partnership. You work with the Orchestrate tech lead, our product manager, and other engineering teams on priorities, customer needs, and the roadmap.
What You'll Accomplish
- First 3 months: Learn the platform and who depends on it. Take over one well-framed project and deliver it using an agent-driven workflow that others can learn from.
- By month 6: Own a larger open problem and bring the team a researched recommendation.
- By month 12: Have shipped at least one major distributed-computing capability to production, and have made the way you work with agents the way the team works.
Your impact is measured by the effect of what you chose to build, as well as by how dependably you deliver it.
What You Bring
Required
- 8+ years of professional software engineering as a hands-on builder, with a record of owning large projects from idea to production.
- A working understanding of distributed systems, shown in work you describe rather than in a skills list: database transactions, replication, consistency, fault tolerance, scheduling, queues, and idempotency.
- Experience building or extending distributed systems yourself, not only operating or consuming them (writing workflow DAGs alone doesn't count). Any language works: Go, Rust, Java, Python, C++, or similar.
- Experience running a software project end to end: breaking it down, sequencing it, and allocating resources, making sound technical bets, and running a quality and review process, whether the work is done by people, agents, or both.
- Hands-on experience using AI coding agents on real work, and a clear view of how you review, test, and trust what they produce.
- Comfort with ambiguity and risk. You can research an open problem, make a recommendation under incomplete information, and stand behind the outcome.
- Clear written and spoken communication, including the ability to explain your reasoning to engineers, product, and customers.
Valued
- Experience moving a monolith or long-running tasks toward a distributed, containerized, Kubernetes-based design
- Working with customers who bring their own compute
- C++ experience (a bonus, not a requirement)
- High-performance computing, GPU workloads, or scientific computing
- CAD/CAM, simulation, or aerospace and manufacturing software
You might not be the right fit if
- You want to manage a team of people. This role leads agents, not direct reports.
- You want every task defined for you. Here you decide what gets worked on.
- You prefer to hand off the hands-on work. We want someone who still builds.