Staff AI Infrastructure Engineer

ABOUT THE TEAM

The Air Dominance & Strike team at Anduril develops aerial and multi-domain robotic systems. The team is responsible for taking products like Fury (unmanned fighter jet) and Barracuda (air-breathing cruise missile) from concept to product. The team also develops Lattice for Mission Autonomy, Anduril’s premier software platform that enables masses of Fury, Barracuda, and other first and third party robots to collaborate across various missions. We work in close coordination with specialist teams like Perception, Motion Planning, Hardware, and Test Engineering to solve some of the hardest problems facing our customers. We are looking for software engineers and roboticists excited about creating a powerful autonomy software stack that includes computer vision, motion planning, SLAM, controls, estimation, and secure communications.

ABOUT THE JOB

We are looking for a founding Staff AI Infrastructure Engineer to architect, build, and scale the end-to-end machine learning platform that powers Anduril’s autonomous systems. 

As a Staff Engineer, you will own the technical roadmap for our ML platform. You will build the robust infrastructure, MLOps tooling, and systems architecture required to train, evaluate, host, and serve complex AI models (including LLMs, computer vision, and RL agents) in both cloud environments and air-gapped, offline tactical edge networks. You will be a force multiplier for our AI Research Scientists, optimizing their experimentation velocity and managing the lifecycle of terabytes of multi-modal sensor and simulation data. Over time, you will help recruit, mentor, and expand this infrastructure engineering team.

WHAT YOU’LL DO

• Design, build, and maintain our foundational training, orchestration, and experimentation infrastructure to support state-of-the-art model development.
• Actively identify, measure, and eliminate bottlenecks in the ML research lifecycle. Build highly automated tools for hyperparameter tuning, model profiling, and experimentation tracking.
• Design and scale robust, high-performance ETL pipelines capable of processing terabytes of multi-modal data (video, camera feeds, radar, flight telemetry, and simulation logs) captured from physical assets and test sites.
• Architect high-throughput, low-latency model serving frameworks optimized for both scalable cloud environments and air-gapped, resource-constrained tactical edge environments. Build CI/CD pipelines for ML models with automated validation, canary deployments, and rollback capabilities.
• Build robust, automated pipelines for continuous evaluation, model validation, and reinforcement learning alignment loops (RLHF/DPO) to guarantee model safety and predictability in high-stakes environments.
• Work closely with AI Researchers, Computer Vision teams, and platform engineers to design unified infrastructure standards across the company's autonomous systems programs.

REQUIRED QUALIFICATIONS

• 7+ years of software engineering experience with a proven track record of designing, building, and operating production-scale machine learning systems and platforms (MLOps).
• Proficient in Python, Go, C++, or similar backend languages. Deep understanding of ML systems design, memory management, and distributed computing.
• Deep experience with containerized deployments (Docker, Kubernetes), GPU scheduling/orchestration, and distributed training frameworks (e.g., PyTorch Distributed, Ray, Slurm, or Megatron-LM).
• Hands-on experience building distributed data pipelines (ETL) and managing massive datasets (terabytes of unstructured/multi-modal sensor data).
• Experience setting technical direction, leading complex system migrations, and mentoring senior engineers.
• Eligible to obtain and maintain an active U.S. Top Secret security clearance.

PREFERRED QUALIFICATIONS

• Experience building and running ML infrastructure, model serving, or software registries within secure, air-gapped, or highly regulated environments (e.g., IL5/IL6, GovCloud).
• Experience specifically building training and evaluation platforms for Large Language Models, Generative AI architectures, or Reinforcement Learning (RL) pipelines.
• Experience profiling training hardware performance, identifying bottlenecks across networks and memory, and optimizing hardware utilization.
• Experience designing and operating multi-tenant ML platforms that serve multiple research teams, with robust resource isolation, quota management, and fair scheduling across shared GPU clusters.
• Hands-on experience with next-generation AI accelerators beyond standard GPUs (e.g., AWS Trainium, Google TPUs, or custom ASICs) for training and inference workloads.
• Experience building production monitoring and observability systems for ML models, including prediction drift detection, data quality monitoring, and automated retraining triggers.