Anduril Industries is a defense technology company with a mission to transform U.S. and allied military capabilities with advanced technology. By bringing the expertise, technology, and business model of the 21st century’s most innovative companies to the defense industry, Anduril is changing how military systems are designed, built and sold. Anduril’s family of systems is powered by Lattice OS, an AI-powered operating system that turns thousands of data streams into a realtime, 3D command and control center. As the world enters an era of strategic competition, Anduril is committed to bringing cutting-edge autonomy, AI, computer vision, sensor fusion, and networking technology to the military in months, not years. 

About the Team: 
Air Dominance & Strike designs, builds, and flies autonomous air vehicles—from  collaborative combat aircraft to expendable cruise missiles and counter-UAS  interceptors. Our vehicles move from whiteboard to first flight on timelines that  traditional primes consider impossible, which means our design cycles live or die  on how fast we can close the iteration loop. The Anduril AI Engineering team  exists to collapse that loop.  

We are engineers first. We work from engineering first principles and unlock capability through machine learning and AI. We are building to scale across CFD, FEA, thermal, and electromagnetics, with pipelines, architectures, and validation practices that carry across programs. 

 About the Job 
We are looking for a Machine Learning Engineer to apply the latest research in  physics ML to the toughest bottlenecks in our design cycle. This role owns the  entire surrogate modeling stack for Air Dominance & Strike—the architectures,  the training infrastructure, the simulation data pipelines that feed it, and the  tooling design engineers use to consume predictions.  

 You will develop, train, and deploy surrogate models that accelerate the physics simulations underpinning our air vehicle programs. Working alongside aerodynamicists, structures engineers, and thermal engineers, your models will directly inform decisions on hardware that actually flies. Where current methods fall short, you will develop new ones, with ample room to identify novel applications of physics ML across our portfolio.  

 Defense experience is not required. We are looking for engineers who came to machine learning through the complex physical problems they were already trying to solve.  

 This role is based onsite in our Costa Mesa, CA office. 

 What You'll Do 

• Own the Surrogate Modeling Stack: Drive the end-to-end design, training, and deployment of production-grade surrogate models to accelerate critical simulation workflows (CFD, FEA, thermal, structural, and aeroelastic) across air vehicle design. 
• Develop State-of-the-Art Architectures: Design and implement neural architectures tailored to engineering physics, developing new techniques for uncertainty quantification, active learning, and inverse problems (such as geometry and shape optimization). 
• Build Robust Data & Training Infrastructure: Create the pipelines behind the training—extracting, aggregating, and sanitizing tens of thousands of high-fidelity results from solver outputs. 
• Optimize & Integrate: Optimize inference for the design loop (maximizing GPU utilization, batched evaluation, and interactive-speed latency) and seamlessly integrate surrogate predictions into the tooling our domain engineers already use. 
• Collaborate & Mentor: Partner with domain engineers to identify where ML delivers the highest leverage, stay current with Physics AI research, and provide technical mentorship to non ML engineers. 

Qualifications 

• Education: BS, MS, or PhD in aerospace, thermal, mechanical, or electrical engineering, or in machine learning/AI/data science with a demonstrated engineering foundation. 
• Experience: 3+ years of experience taking ML models from R&D into production using large-scale scientific or engineering datasets. 
• Physics ML Expertise: Working knowledge of modern surrogate architectures (e.g. GNNs, Transolver, DoMINO & GeoTransolver) combined with hands-on experience running physical simulations (CFD, FEA, thermal, etc.) and a command of the underlying numerical methods.  
• Software & Frameworks: Proficiency in Python and MATLAB; experience with PyTorch, TensorFlow, and NVIDIA PhysicsNeMo (Modulus); and experience developing on Linux with GPU accelerators and distributed training. 
• Data & Engineering Best Practices: Track record of building production data pipelines from heterogeneous engineering sources, utilizing uncertainty quantification, conducting statistical analysis, and building data science dashboards  
• Clearance: Must be a U.S. Person eligible to obtain and maintain a U.S. Top Secret security clearance  

Preferred Qualifications 

• Advanced Physics ML: Graduate research focused on AI for scientific simulation, experience solving inverse problems (geometry optimization/design under uncertainty), and hands-on experience building active learning or adaptive sampling pipelines. 
• Domain Expertise: Prior work in aerospace, automotive, turbomachinery, or another simulation-heavy hardware domain, with familiarity in commercial solvers, meshing tools, and CAD interoperability.
• Advanced Tooling: Working knowledge of foundational ML methods (Gaussian processes, XGBoost, Elastic Net regression & clustering) with the ability to build custom architectures, advanced skills in visualization software (Plotly, Seaborn, Matplotlib), and ML Ops orchestration experience (e.g. Docker, Weights & Biases, AWS S3, Lambda & SageMaker)