Software Engineer

Software Engineer — Distributed Systems & High-Performance Computing Location: New York City, USA (on-site) Employment type: Full-time Compensation: $200,000 – $300,000 + equity About the role We are recruiting on behalf of a frontier AI lab building models that reason about the physical world — real-time inference over complex sensor data, in defense and security settings. The company is headquartered in New York, recently funded, and founded by researchers and engineers from leading frontier model labs and quantitative trading firms. The team is small and growing, and this hire will work directly on the systems that make their models run in the field. The client's identity is shared with candidates during the first call. The role We are looking for engineers interested in high-performance systems, distributed infrastructure, and deployment in constrained environments. The model isn't the hard part here. The hard part is everything around it: running inference inside a power and memory budget, on hardware you don't control, in environments where the network is unreliable and there's no quick path to ship a fix. Ideal candidates have experience building reliable and scalable systems in research or production environments. Areas of work include: Distributed systems and compute infrastructure Edge deployment and runtime optimization High-performance data processing Systems reliability and observability Infrastructure tooling and automation Performance-critical software systems You'll likely be a fit if you have: Built and operated systems where latency, throughput or resource limits were the primary design constraint Profiled and optimized real workloads rather than reasoning about performance in the abstract Worked in a systems language (Rust, C++, Go) alongside Python on AWS Owned what you shipped in production — deployment, monitoring, incident response Comfort with ambiguity and a small team where scope isn't handed to you How to apply Apply through this posting. Momi US manages the process end to end; you'll speak with me before any client introduction.