Anduril Industries is a defense technology company aiming to transform military capabilities by integrating advanced AI, autonomy, computer vision, sensor fusion, and networking technology. The Frontier AI team specifically focuses on productizing cutting-edge AI innovations like LLMs for national security applications, addressing infrastructure challenges for deploying these models in edge and air-gapped environments.
Requirements
- Comfortable scripting or coding in Python and/or Golang and have experience integrating models with external systems or APIs
- At least 3 years of experience creating and deploying production level software
- At least 2 years of experience deploying and integrating LLMs into production software
- Experience with modern build and deployment tooling such as Nix and Terraform
- Experience with DOD Simulators (AFSIM, NGTS, etc.)
Responsibilities
- Own the software solutions that are deployed to customers
- Lead the discovery, development, and implementation of workflows and tools for training and adapting LLMs to generate simulations
- Build and maintain test suites for evaluating LLM interactions with simulation environments (e.g., automated correctness and performance testing)
- Design, develop, and maintain evaluation pipelines for geospatially aware LLMs, including customized evals for spatial reasoning and location capability
- Contribute to ongoing efforts to fine-tune LLMs for improved performance in simulation script generation and multi-modal input handling
- Document processes, findings, and results for internal knowledge transfer and external reporting
Other
- Eligible to obtain and maintain an active U.S. Top Secret security clearance
- Experience working in small teams such as a high-growth startup
- Ability to rapidly adapt and perform as a generalist when needed
- Passionate about working on problems that have a real-world impact
