Staff Software Engineer, Gustie Assist
Core
Design and build an LLM agent runtime and full stack for Gustie Assist (Otto), enabling support advocates to autonomously reason and take actions on customer payroll, benefits, and compliance issues.
Role type
Staff Software Engineer (Applied AI / Agentic Systems)
Builds
Otto, the AI agent platform that performs real actions across Gusto's systems for small business customers
Domain
Fintech / Small Business HR / Applied AI
Deliverable
production ML models | product features
Required skills
End-to-end system ownership, LLM-powered agentic systems, tool-calling agents, retrieval, evals, full-stack development, safety and blast radius judgment, mentorship
Preferred skills
Agent frameworks, LLM-as-judge, MCP-style tool architectures, Ruby on Rails, JavaScript/React, GraphQL
Technologies
LLM runtime, Rails, GraphQL, React
Responsibilities
Design and build the Otto platform runtime, agent graph, and tool/action layer; Expand agent capabilities for autonomous actions with human-in-the-loop; Build eval and observability systems for production monitoring; Work directly with LLMs on prompting, tool design, and feedback loops; Partner with Product, Data, Design, and CX teams; Mentor engineers on building agentic systems
Seniority
Staff, hands-on IC with mentorship