Engineering Intern

Not a resume internship. Own a scoped project end to end, ship it to production, and demo weekly. Real product, real stakes.

Large companies run critical operations on complex enterprise systems. These environments are difficult to work with, and mistakes are expensive.

We are building AI systems designed to operate safely inside these environments. That means understanding real system behavior, handling imperfect data, and making sure outputs can be trusted before anything is acted on.

MCP is part of how we do that. We build MCP-based capabilities and agent workflows that connect AI to real systems in a controlled way.

Small team, real product, real stakes.

This is not a resume internship. You will be expected to ship real work.

You will work with senior engineers on a scoped project, own it end to end, and ship it to production. The exact project depends on what we need during your term, but examples include:

• Build or improve an MCP-based capability for the system to inspect and reason about part of an enterprise environment
• Implement backend logic for detecting and surfacing business issues
• Design a product surface that shows findings clearly and supports user action
• Write tests, validation logic, and regression checks for features we are preparing to ship
• Document bugs, edge cases, gaps, and improvement ideas clearly as the system evolves

You will open PRs, go through code review, iterate, and deploy to production. You will demo your work weekly.

You will be expected to think about correctness, edge cases, and what happens when the system is wrong, incomplete, or ambiguous.

You will work closely with the senior team, including the founder. The team is small enough that your work will directly impact the product, and you will see it used before your term ends.

• Enrolled in computer science, software engineering, or a similar program
• Something you have built outside of class that you can show and explain
• Curiosity about AI-assisted development

Bonus if you already use tools such as Claude Code, Cursor, or Git Hub Copilot. Familiarity with MCP, testing, or agent workflows is also a plus, but not required. Enterprise software exposure is great to have, not required.

• What production systems look like under real constraints
• Code review, testing, regression validation, and deployment in a working product
• How to use AI coding tools with judgment
• How MCP-based capabilities and real enterprise workflows get built in practice

Competitive market rate. Full-time conversion possible.