PRS Facility Location
EMD HQ LaGrange, IL - 400
Job Purpose
As an AI Intern on the ART team, you'll help design and prototype AI strategies that interpret complex rail logistics scenarios and recommend (and where appropriate, automatically execute) planning actions on top of our optimization and decision support tools. You'll work at the intersection of large language models (LLMs), decision science, and software engineering-building integrations that let LLMs reason over domain data, call specialized tools and services, and generate reliable, auditable outputs for planners and operators. You'll collaborate closely with product, R&D, data engineering, and optimization teams to deliver experiments, proofs of concept, and pilot features that improve network velocity, asset utilization, and safety.
Req ID
11428BR
Company Description
Progress Rail stands at the intersection of legacy and innovation-driving the future of rail with a pioneering spirit. Since its founding in 1983, the company has grown to become one of the world's largest and most trusted providers of railroad products, services, and technologies. Today, Progress Rail delivers a comprehensive portfolio of cost-effective solutions to railway customers around the globe. From the rails themselves to the EMD® locomotives that ride them, the company's products are in operation across more than 100 countries-powering progress and connecting communities. In 2006, Progress Rail joined Caterpillar Inc., further strengthening its ability to lead the rail industry with cutting-edge technology, unmatched expertise, and a commitment to excellence. At Progress Rail, the team is not just building the future of rail-they are making history every day.
Education / Training
• Currently enrolled in a Bachelor's or master's program in Computer Science, Data Science, Electrical/Computer Engineering, Operations Research, Industrial Engineering, Applied Math, or a closely related field.
• Coursework in at least two of the following: machine learning, natural language processing, optimization/operations research, algorithms, databases, data structures, probability/statistics, or software engineering.
Job Title
AI Intern, Advanced Rail Technology
City
La Grange
Key Job Elements
• Prototype LLM-enabled workflows: Build agents, prompts, and tool-use policies that let LLMs interpret timetables, work orders, speed restrictions, and plan alternatives, then suggest actions aligned with operating constraints.
• Integrate with optimization & decision tools: Connect LLM pipelines to solvers, simulators, and rules engines (e.g., scheduling/dispatch optimization, what if simulators) and validate that recommendations respect operational and safety rules.
• Engineering hygiene: Package prototypes as services/notebooks with versioned prompts, config, and unit tests; document assumptions, data contracts, and model/solver interfaces.
• Collaboration: Work with software, optimization, and R&D partners to iterate quickly on requirements; demo progress to stakeholders and capture feedback for the next sprint.
• Compliance & safety: Apply data privacy, access controls, and safety guardrails; follow MLOps and software best practices for deployment-readiness in an industrial environment.
Region
United States
Qualifications and Experience
Required:
• Hands on experience (course, project, or internship) with Python for data/ML (e.g., pandas, NumPy) and at least one ML/NLP library (e.g., PyTorch, TensorFlow, scikit learn, Hugging Face).
• Basic understanding of LLMs (prompting, fine tuning/LoRA, RAG, function/tool calling) and how to evaluate them (accuracy, grounding, latency, cost).
• Familiarity with REST APIs, JSON, and microservice integration; ability to consume/produce structured tool outputs.
• Strong problem-solving skills, clear written communication, and the ability to collaborate in a fast paced, multidisciplinary environment.
Preferred:
• Exposure to optimization/OR (linear/integer programming, heuristics, constraint programming) and integrating AI with solvers/simulators.
• Experience building RAG pipelines (vector stores, chunking, embeddings), prompt versioning, and guardrails/red teaming.
• Knowledge of MLOps practices (experimentation tracking, model/data versioning, CI/CD for ML) and containerization (Docker).
• Domain interest in transportation, rail logistics, supply chain, or cyber physical systems
Essential and Physical Activities Functions:
• Strength - This is a traditional office setting and requires the occasio