Role description
DunedainAI/ML Engineer (Full-Time)
Dunedain | Austin, TX
ABOUT US
Dunedain is building the battlefield operating system of the future and fielding it today.
We turn the digital noise of combat and training into a single coherent picture. Our products let operators plan, decide, and act at machine speed, in the fight alongside the autonomous systems maneuvering beside them. Warfighting formations that run our stack gain a living intelligence, one that learns how they fight and grows with them, from fireteam to Corps.
Unlike most military technology vendors, we are not a concept, a slide deck, or hot air. We are not waiting to be told what to build for the future fight. We're a team of former Warfighters and elite developers, scaling aggressively to give the US military and allies the most unfair decision and tempo advantage in human history.
ROLE OVERVIEW
As an AI/ML Engineer, you'll design and optimize the agentic retrieval pipelines that let our agents pull the right military data and reason over it in context. That means agentic RAG and agentic graph RAG, LoRA fine-tuning of language models for defense use cases, and the vector and graph infrastructure behind them. You'll work with our engineering and product teams to ship AI that improves decisions and cuts the busy work out of defense workflows. Full-time, hybrid in Austin, TX, with travel up to 25%.
KEY RESPONSIBILITIES
Design, build, and optimize agentic RAG and agentic graph RAG pipelines for military data retrieval and contextual generation.
Fine-tune language models for defense applications using LoRA and related methods, so technical data and assets are easier to reach and use.
Build the retrieval backbone: vector databases, graph stores, and indexing strategies that stay fast and relevant at scale.
Experiment with NLP, NLU, and NLG techniques to push the capability and performance of our AI products.
Work with software engineers to get models into production, with scalability, reliability, and operational use demands.
Partner with engineering and product to fold ML capabilities into the platform end users actually touch.
Provide technical leadership across the product development process.
REQUIRED SKILLS AND EXPERIENCE
Agentic RAG & NLP: Hands-on experience with RAG systems, ideally agentic and graph-based, plus vector databases and language model fine-tuning. Solid grasp of NLP fundamentals underneath the pipeline.
Fine-tuning: Practical experience adapting language models with LoRA or similar parameter-efficient methods.
Programming: Strong Python and real proficiency with PyTorch or TensorFlow.
Production experience: You've shipped ML into production on a cloud platform (AWS, GCP, or Azure) with containerization (Docker, Kubernetes), working with software engineers to integrate models into live applications.
Problem-solving: You can drive a project from concept to execution on your own and keep it aligned with the mission and the needs of defense end-users.
US Citizenship required, and willingness and eligibility to obtain a US security clearance
PREFERRED QUALIFICATIONS
Experience with agentic graph RAG or knowledge-graph-backed retrieval.
Military experience.
Startup experience, or a track record in fast-moving, agile projects.
Familiarity with the defense sector, including the data security and compliance requirements that come with it.
Experience building for government or defense users, where the specific workflow matters.
BENEFITS
A direct hand in bringing AI to the defense mission, on a small, fast-moving team.
Hybrid in Austin, TX or Remote
Competitive total compensation of $150,000 - $300,000+ 0.1-0.5% company equity, based on experience.
TO APPLY
Send your resume and a short cover letter. Tell us about your work with agentic RAG, fine-tuning, and NLP, how you solve hard problems, and why you want to build this at Dunedain.