Full Stack Engineering Intern

Nash•San Francisco, CA

5d•Hybrid

About The Position

We're looking for a Full Stack Engineering Intern to work on Nash's intelligent and agentic layer: the part of the system that senses conditions on the ground, decides what should happen next, and acts. Agents that triage exceptions, recover routes mid-shift, answer operator questions, and intervene before a promise breaks. You'll work across React and Python to build the interfaces operators trust and the agents working alongside them. This isn't a sandbox internship. You'll pick up real work on day one, ship code customers use, and leave with something concrete to point to. The right person here is as interested in how an agent reasons as in how its decisions show up in the interface. Agents are only as useful as the trust operators place in them, and that trust is built or broken in the product.

Requirements

Currently pursuing a degree in Computer Science or a related field, or recently graduated.
Hands-on experience building full-stack web applications, through coursework, side projects, prior internships, or open source.
Working knowledge of Python, React (or an equivalent modern framework), TypeScript, HTML, and CSS.
Familiarity with REST APIs and relational databases (PostgreSQL preferred). Exposure to GraphQL or NoSQL is a plus.
Curiosity about LLMs, agents, and AI systems. Prior experience building with them is a plus, not a requirement.
Product sense and craft. You care about how things feel, not just whether they work.
A strong communicator who can collaborate across disciplines and drive ambiguous problems toward a clear outcome.
High agency. You don't wait to be told what to do next.
Strong preference for the Bay Area. Nash runs a hybrid-first product team out of our San Francisco office.

Nice To Haves

Exposure to GraphQL or NoSQL is a plus.
Prior experience building with LLMs, agents, and AI systems is a plus, not a requirement.

Responsibilities

Ship features end to end across Nash's agentic layer, from the React and TypeScript surfaces where operators work with agents, to the Python services and LLM workflows that power them.
Build agent capabilities that resolve real operational problems: rerouting around a closure, reassigning a stuck order, surfacing the right answer to a dispatcher mid-shift.
Translate logistics problems into product. Sit in with PM, design, and operations to turn messy real-world workflows into systems that fit how customers actually run their business.
Contribute to system design discussions on how agents reason, when they act on their own, and when they hand off to a human, with senior engineers around the table to push your thinking.
Write clean, well-tested code, participate in code reviews, and pick up the standards a strong engineering team runs on.
Debug across frontend, backend, and model behavior, with the observability and evals that catch problems before customers do.
Build responsive interfaces with a sharp eye for latency, loading and error states, and the edge cases that decide whether an operator trusts an agent's recommendation.