Agent Engineering Intern

Cheiron•Los Altos, CA

8d•$35 - $55•Onsite

About The Position

We're looking for an Agent Engineering Intern to build new agent-driven workflows inside Cheiron's product. Cheiron's agents do more than generate an answer. They break complex objectives into steps, retrieve evidence, call tools, maintain state across long-running work, recover from failures, and produce structured results that a person can review and trust. You will own a bounded workflow and help take it from an ambiguous problem to a reliable, working product. This is an applied engineering role, not an AI research internship. We care about strong software fundamentals, hands-on ownership, and evidence that you have built nontrivial systems yourself. The role is best suited to undergraduate and master's students who are already serious builders.

Requirements

Currently pursuing an undergraduate or master's degree in computer science, computer engineering, or a related technical field.
Strong programming fundamentals and experience building a substantial software project, developer tool, backend system, or open-source contribution.
Hands-on experience building a nontrivial agent, retrieval system, tool-using workflow, evaluation harness, or multi-step automation system.
Clear personal ownership of the work: you can explain what you implemented, the decisions you made, and what you would improve.
Ability to reason about state, failure modes, grounding, evaluation, and engineering tradeoffs—not only prompts and model selection.
High agency and comfort working through ambiguous problems in a fast-moving environment.
Ability to work fully in person in Los Altos. Semester or quarter interns must commit at least 20 hours per week; summer interns should be available full time.

Nice To Haves

Experience with Python, FastAPI, Postgres, agent frameworks, retrieval systems, or vector databases.
Backend, distributed-systems, infrastructure, data-pipeline, or end-to-end product experience.
Experience with tests, deployment, observability, security, or reliable failure handling.
Experience building systems involving citations, provenance, human review, or other high-stakes correctness requirements.
Strong open-source work, prior software-engineering internships, or an independently shipped product with real users.

Responsibilities

Design and implement new agent workflows with explicit state, task decomposition, tool use, and typed outputs.
Build retrieval and citation paths that ground important outputs in source evidence.
Add evaluation, tracing, retries, observability, and recovery behavior for multi-step execution.
Build the backend services, data models, APIs, and focused product surfaces required to make the workflow useful.
Test agent behavior against realistic success cases, failure cases, and incomplete evidence.
Work directly with engineers and domain experts, demonstrate progress frequently, and make pragmatic scope decisions.