Graduate AI Engineer Intern Summer 2027

MiniMed•Johns Creek, GA

25d•Onsite

About The Position

Join MiniMed for a summer internship where you’ll do meaningful work, gain hands-on experience, and make an impact on our mission to make every day a better day for people with diabetes. You’ll be part of a collaborative intern community, build your professional network, and gain exposure to opportunities that help you grow both personally and professionally. Our summer internships are full-time (40 hours/week), paid positions lasting 10-11 weeks. We offer competitive compensation, educational and social programming throughout the summer, as well as housing assistance and relocation support for eligible interns. At MiniMed, we're transforming how a medical device company operates through artificial intelligence, and we want you to help drive that transformation. As an AI Engineer Intern, you'll be embedded within MiniMed's AI & Data Science team, working at the intersection of cutting-edge AI technology and real-world business impact. This is a hands-on, build-and-ship role, not a research or observation role. You'll own a scoped project from problem definition to prototype development and evaluation, partnering with technical teams and business stakeholders to deliver measurable business value. Your work will span two of our core AI focus areas: AI Agents and Tools: Design, build, and evaluate LLM-powered agents on our MIA stack. You’ll work with industry-standard technologies, including AWS Services, Databricks and LangGraph, to develop reliable solutions for real business needs. Enterprise AI Enablement: Help expand self-service AI adoption across the enterprise by building and deploying reusable agents and function-specific skills. You’ll empower teams to apply AI in their day-to-day workflows and improve how they work. Project examples may include: Prototyping a domain-specific AI agent for a business function Developing and running evaluation test sets to measure agent quality, detect hallucinations, and identify regressions Building reusable agent skills and applying guardrails to ensure responsible, auditable outputs Along the way, you'll gain practical grounding in evaluation-driven development, exposure to responsible AI delivery in a regulated medical-device environment (privacy, auditability, human-in-the-loop design), and real-world experience translating a business need into a shipped, measurable solution. You'll leave with a strong portfolio of applied AI work and a deeper understanding of how enterprise AI strategy gets executed in a complex, compliance-driven industry.

Requirements

Working toward a Master’s degree in Computer Science, Data Science, Machine Learning, AI, or a related discipline with an anticipated graduation date of Winter 2027 or Spring 2028.
Currently authorized to work in the United States.
Applicants must not require employment sponsorship now or in the future.

Nice To Haves

Strong Python programming skills
Familiarity with AI/ML concepts including; LLMs, prompt engineering, RAG, deep agents and model evaluation
Exposure to tools such as Databricks, Spark, LangGraph, LangSmith, vector databases, or SQL
Coursework or project experience in machine learning, NLP, deep learning, statistics, and algorithms
Strong problem-solving skills, curiosity, and the ability to navigate ambiguous, real-world challenges
Effective communication and collaboration skills

Responsibilities

Design, build, and evaluate LLM-powered agents on our MIA stack using industry-standard technologies, including AWS Services, Databricks and LangGraph.
Develop reliable solutions for real business needs.
Expand self-service AI adoption across the enterprise by building and deploying reusable agents and function-specific skills.
Empower teams to apply AI in their day-to-day workflows and improve how they work.
Prototype a domain-specific AI agent for a business function.
Develop and run evaluation test sets to measure agent quality, detect hallucinations, and identify regressions.
Build reusable agent skills and apply guardrails to ensure responsible, auditable outputs.