Research Intern - Agentic Data Memory

What you'll do

• Work on a defined research project with mentor guidance and collaborate with interns and researchers.
• Advance research in areas such as graph-based methods, agent memory and data, workflows, and evaluation.
• Present findings and contribute to a final presentation; publication is a possible goal.

What they ask for

• Currently enrolled in a doctorate degree program in Computer Science or a related STEM field.
• Physically located at the manager’s Microsoft worksite for the internship.
• Submit at least two reference letters, a cover letter, and relevant work or research samples.

Nice to have

• At least 1 year of research experience in a relevant area, demonstrated through publications or preprints; experience with LLM-based agents, RAG, knowledge graphs, or agent memory systems is strongly preferred.
• At least 1 year of Python experience and at least 1 year in Calculus, Probability and Statistics, and/or Machine Learning; benchmark design, statistical evaluation, and GPU or cloud experience are additional pluses.

Pay

USD $6,810 - $13,480 per month (base salary); USD $8,920 - $14,610 per month (base salary, San Francisco Bay area and New York City metropolitan area)

Tools

Python, PyTorch, GPU clusters, cloud infrastructure, LLM-based agents, retrieval-augmented generation (RAG), knowledge graphs

AI focus

Research on artificial intelligence and agentic systems, including graph-based methods, agent memory and data, workflows, harnesses, loop engineering, and evaluation.