What are the responsibilities and job description for the ML/NLP Research Engineering Intern position at Dunedain?
ML/NLP Research Engineering Intern
Dunedain Systems | Hybrid (Austin, TX or Pittsburgh, PA)| Summer 2026 Internship
May 26 - August 15, 2026 (12 weeks) | Project Completion Bonus
About Dunedain
Dunedain builds AI-powered planning and decision tools for military units. Our platform combines frontier language models, retrieval systems, and agentic workflows into mission-ready software that works in disconnected, classified environments. We're a veteran-founded startup backed by Tier 1 venture capital, with active pilots across the military.
The Role
We're looking for a motivated ML/NLP Research Engineering Intern to join our founding team for Summer 2026. You'll bring expertise and state-of-the-art research into the system, underpinning the tools and infrastructure that our AI agents rely on every day. You will be contributing to advancing the Deep Learning, Machine Learning, and NLP capabilities of our models through novel architectures, designs, and experimentation.
This isn't a sandbox internship. This is graduate-level research designed to have world-class impact. The research you do and models you create will ship to operational military units. You'll work directly alongside our founding engineers, own real projects from day one, and leave the summer with research production experience that most interns don't get until years into their career. You will also have the opportunity to present findings to the team, communicating the scientific process, experiments, and differentials.
What You'll Do
- Design and build ML Models serving as the ontological layer of an Agentic Systems
- Conduct research staying on top of latest papers, metrics, approaches to our problems
- Define Metrics, Benchmarks, and curate dataset for testing models you design
- Participate in R&D experiments to prototype new models and their capabilities
- Document your research project and insights so the team and the agents can use them effectively
- Own model development end-to-end with mentorship from senior engineers
Research Areas
- Long-context NER and Ontology Generation from multi-document corpora
- Structured reasoning over noisy, non-standard real-world text
- Evaluation of agent reliability of constrained environments
- Hybrid symbolic and neural representations
- Spatio-Temporal Graphical representations
You'll be paired with a senior engineer mentor and have full access to our tech stack from day one. Your main focus is building the ML/NLP models our agents rely on, not ETL pipelines. You will be provided with clearly documented target outcomes, constraints, and impact statements.
Qualifications
- Currently enrolled (rising senior, master's, or recent graduate) in Computer Science, Data Science, Engineering, Statistics, Linguistics, Cognitive Science or related field.
- Familiarity with ML/NLP concepts: Deep Learning for NLP, NER, Symbolic/Neural Representations, Multivariate Calculus, Syntax Semantics
- Strong proficiency in Python, Familiarity with PyTorch
- Documentation, communication, presentation of scientific approaches to ML
- Driven by the mission: you want your work to reach the people who need it most
- U.S. Person status is required as this position needs to access export-controlled data
Nice-to-Haves
- Experience building NLP/DL/ML models
- Prior research involving ML/CL/NLP Experimentation
- Version Control (git), Cloud Systems (AWS, GCP, Azure)
- Active security clearance (Secret or TS-SCI)
- Familiarity with Pydantic AI, LangChain, or similar agent frameworks
Location and Schedule
- Remote Internship, with option for Hybrid in Austin, TX (working out of Gevity, a high quality coworking space)
- Full-time hours during the 12-week program
- Possibility of part-time extension after August based on performance
What We Provide
- Project completion bonus
- Cursor / GitHub Copilot subscription for the duration of the internship
- Three-month membership to Gevity (gym, sauna, ice baths, coworking)
- Direct mentorship from founding engineers translating latest research into value for operational military units
- Real ownership of research project and model development that
- High-impact work with a direct line to the mission
- Hybrid flexibility and a fast-moving, veteran-led team
Dunedain is an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability, or veteran status.