AI Agent Memory Intern

DataRobot; Massachusetts; On-site; USD 28,000 - 41,000; Full time

Job summary

DataRobot is offering a hands-on internship in a fast-paced AI environment in the United States. You will work on MCP servers and memory, connecting agents to tools, shaping context, and designing what an agent remembers and forgets.

You will write code, experiment with ideas, observe outcomes, and ship improvements rapidly. A strong candidate enjoys learning, communicates clearly, and proactively helps teammates unblock their work.

Qualifications

• Comfort writing and reading Python; eager to improve quickly.
• Curiosity about how AI agents work: tools, prompts, context windows, and why an agent succeeds or fails on a real task.
• Interest in MCP — servers exposing tools, simple, reliable contracts.
• Interest in memory for agents: what to store, retrieve, and how it guides next steps.
• Ability to take a vague problem, try a small version, and show what you learned.
• Clear communication: say when stuck, share attempts, help others move faster.
• Good engineering habits: readable code, small tests, and notes for future maintainers.
• Open to feedback and adapting approaches when asked.

Responsibilities

• Building and improving MCP servers and clients: tool definitions, inputs/outputs, error handling, and runnable examples.
• Experimenting with agent memory: what gets saved, how it is retrieved, and its effect on task quality.
• Wiring tools and memory into an agent loop and observing where it stalls or errs.
• Making small, clear changes that unblock teammates—docs, tests, fixes that reduce friction.
• Turning rough ideas into runnable code and tightening based on results.
• Pairing with people who know the system well, then taking ownership of the next slice.

Skills

Python

AI agents

MCP server

Memory design

Problem solving

Communication

Readable code

Testing

Feedback

Tools

Agent frameworks

Tool calling

Job description

DataRobot is offering a hands-on internship in a fast-paced AI environment in the United States. You will work on MCP servers and memory, connecting agents to tools, shaping context, and designing what an agent remembers and forgets.

You will write code, experiment with ideas, observe outcomes, and ship improvements rapidly. A strong candidate enjoys learning, communicates clearly, and proactively helps teammates unblock their work.