Advanced Software Engineer - Ai Surfaces

Relativity•Illinois, IL

3d•$103,000 - $155,000•Hybrid

About The Position

This role offers an opportunity to grow as a full-stack AI engineer on a small team that ships agents to lawyers weekly. You will work alongside the engineers who built Gavel from the ground up and gain direct exposure to the founders. You will begin with well-defined features across the stack and grow toward end-to-end ownership of agentic features. This Advanced Software Engineer role is designed for an early-career to mid-level engineer ready to grow quickly. The role reports to a Lead or Staff Engineer on Gavel at Relativity and may be based remotely in the United States or in Seattle or Chicago.

Requirements

2+ years of professional software development experience or equivalent demonstrated ability, such as a strong open-source track record, a shipped product, or substantial internship experience.
Working proficiency in at least one of TypeScript or Python, plus hands-on experience building full-stack web applications with a front end such as React, services, and a relational database using SQL.
Hands-on experience building an agentic solution with LLMs, even at a small scale, such as an agent, a retrieval-augmented generation application, or a tool-using assistant. Ability to explain system prompts, tool calls, embeddings, and common failure points.
Basic familiarity with cloud services such as AWS, Azure, or GCP, containers, and CI/CD.
Strong fundamentals in data structures, algorithms, and writing clean, tested code.
Curiosity about how lawyers work and the humility to learn from acceptance data when an initial idea does not perform as expected.
Clear written and spoken English and a preference for simple solutions over clever ones.

Responsibilities

Build features across the full stack of Gavel Exec, including agent tools, prompt and context changes, services and API endpoints, and front-end work in the Word add-in and web app.
Write and extend evaluations, run them, and learn to interpret the results.
Ship changes behind feature flags and monitor production outcomes.
Learn and practice a modern software development lifecycle from day one through small, frequent commits, feature-flagged deployments, fast rollbacks, and best-in-class agentic engineering tools.
Fix bugs, improve reliability, and add tests while developing an understanding of how agentic systems fail.
Pair with Senior and Lead engineers on design and code reviews.
Ask questions and document what you learn so others can benefit.
Take on increasingly ambiguous problems as you demonstrate readiness.