Spring 2027 AI/ML Engineering Intern

The Nuclear Company•Washington, DC

18h•Onsite

About The Position

The Nuclear Company is seeking an AI/ML Engineering Intern for Spring 2027. This is a six-month co-op position available from January to June, aligned with the academic calendar. The role is based in Washington DC, requiring on-site work five days a week. The company pioneers a fleet-scale approach to building the next generation of nuclear reactors, utilizing an AI-first mindset where all employees leverage AI, technology, and the Nuclear Operating System (NOS) to enhance their work. NOS is the software that runs regulatory, supply chain, cost, schedule, and field work for real reactor construction, contributing to national security, energy independence, and climate goals. As an AI/ML Engineer Intern on the Platform Integration & AI/Data squad, you will work on machine learning and agent capabilities within NOS, including LLM-driven workflows, retrieval-augmented generation, and agents that interact with NOS data and tools. You will also be involved in the evaluation and monitoring of AI features in production. This role offers the opportunity to work alongside full-time engineers and nuclear domain experts, ship production-ready models and features, and see your work directly applied to building power plants.

Requirements

Currently enrolled in a BS in Computer Science, Machine Learning, Data Science, Statistics, Mathematics, or a related technical field.
Available for a full six-month term and returning to school afterward.
Strong Python fundamentals and hands-on exposure to modern ML and LLM tooling (PyTorch, Hugging Face, or similar) through coursework or projects.
A track record of shipping working software or ML projects (course projects, personal projects, or prior internships or co-ops).
U.S. Person status (U.S. citizen or lawful permanent resident) is required.
Willing and able to work on-site in Washington DC, five days a week, for the full six-month term.

Nice To Haves

Hands-on experience with LLMs and agents: prompting, fine-tuning, retrieval-augmented generation, evals, or tool use (MCP or similar frameworks).
Coursework or project experience with Palantir Foundry and AIP, AWS, or another major cloud platform — Foundry exposure is heavily weighted.
Solid data fundamentals: comfortable doing exploratory data analysis and working with large, messy datasets, including time-series or sensor data.
Exposure to MLOps tooling (MLflow, SageMaker, Vertex AI) or CI/CD for models.
Interest in energy, national security, industrial software, or regulated industries.
Familiarity with nuclear engineering concepts (not required, we teach the domain).

Responsibilities

Ship ML and AI features inside NOS applications teams use every day: site evaluation, red flag analysis, scheduling, lessons learned, and the stakeholder and project tools running across active TNC projects.
Build AI agents and agent workflows that operate against NOS data and tools, including the MCP interfaces and tool-use frameworks they depend on.
Develop data pipelines and integrations across Palantir Foundry, AWS GovCloud, and more — and contribute to the data ontology so models and agents can use NOS data reliably.
Build predictive and anomaly-detection models that support operations and engineering decisions, including time-series analysis of sensor and operational data from active projects.
Help own model quality end-to-end: eval criteria, acceptance thresholds, and regression suites that keep NOS audit-ready and results reproducible; triage issues and root-cause failures.