Staff Gen AI Research Scientist
Anduril Industries · Costa Mesa, California, United States; Seattle, Washington, United States; Washington, District of Columbia, United States
About this role
**Anduril Industries — Staff Gen AI Research Scientist** **About the Team** The Air Dominance & Strike team at Anduril develops aerial and multi-domain robotic systems. The team takes products like Fury (unmanned fighter jet) and Barracuda (air-breathing cruise missile) from concept to product, and builds Lattice for Mission Autonomy—software that enables masses of Fury, Barracuda, and other robots to collaborate across missions. You’ll work closely with Perception, Motion Planning, Hardware, and Test Engineering to solve hard autonomy problems. **About the Job** Anduril is seeking a **founding ML expert** to design, fine-tune, and deploy next-generation **generative AI, LLMs, and agentic systems** for air-dominance platforms and collaborative autonomous behaviors. This is a highly applied research role (~**60% applied research/experimentation** and **40% hands-on coding**) focused on making state-of-the-art LLM models **smaller, faster, and smarter**. You’ll work on both **offboard systems** (mission planning, modeling, simulation) and **onboard systems** (optimizing models to run on power- and compute-constrained edge hardware). You’ll have significant autonomy to set technical direction, define data collection strategy across test sites and simulations, and influence how multi-agent autonomy is deployed in critical missions. **What You’ll Do** - **Build & Fine-Tune Core Models:** Develop, pre-train, and fine-tune in-house LLMs and multimodal foundation models; apply SOTA post-training alignment (SFT, RLHF, DPO) to maximize capability while minimizing cost/footprint. - **Deploy at the Edge:** Architect and optimize models for tactical edge compute; improve latency, memory, and execution speed via quantization, distillation, and pruning. - **Develop Agentic Workflows:** Design agentic architectures, multi-agent coordination frameworks, and planning loops for complex multi-domain missions. - **Drive Multimodal Sensor Integration:** Collaborate with perception/motion planning teams to reason over camera feeds, radar, telemetry, and text-based operational orders. - **Shape Data & Evaluation Strategies:** Define data collection across physical assets, test sites, and simulations; build scalable evaluation frameworks for performance, reliability, and safety. - **Rapid Prototyping to Production:** Prototype with customers, iterate quickly, and scale into production-grade features across the system family. **Required Qualifications** - **Core Engineering Strength:** Production-level coding in **Python** and deep learning frameworks (**PyTorch** or **JAX**). - **ML & LLM Expertise:** Hands-on experience training, fine-tuning, and evaluating **LLMs**, **Generative AI**, or **multimodal models**. - **Solid Foundation:** Background in **Computer Vision, NLP, Robotics, or Speech** with **2+ years** focused on generative models and modern transformer architectures. - **Tooling Familiarity:** Experience with modern training/alignment/orchestration tools (e.g., **Axolotl**, **Hugging Face**, **DeepSpeed**, **Megatron-LM**, **LangChain**, **LlamaIndex**). - **Problem-Solving & Adaptability:** Ability to operate in a fast-paced environment, translating ambiguous mission requirements into concrete code and functional systems.
Listing freshness
CronJobs last confirmed this listing 1d ago. If its source stops confirming the opening for seven days, this page is removed from active inventory.