Staff Software Engineer, AI Developer Tooling

Sentry Sentry · Enterprise · San Francisco, CA · Engineering

Staff Software Engineer focused on AI-assisted coding agents within the Platform Engineering team. The role involves making internal developer systems API-ready for AI agents, building harness tooling and context systems for AI-generated pull requests, and automating engineering tasks. The goal is to integrate AI agents into the software development lifecycle to improve efficiency and quality.

What you'd actually do

  1. Audit Sentry’s internal developer systems and make them API-ready for AI agents. You’ll prioritize and drive the work of exposing those systems programmatically, and build the connections that allow agents to operate on them end-to-end.
  2. Build the harness tooling, context systems, and feedback loops that help agents generate high-quality, repository-aware pull requests, including automated pre-review checks and PR quality measurement tailored to Sentry’s codebase.
  3. Automate high-volume, low-priority engineering work: security dependency upgrades, minor bug fixes, and routine maintenance, so engineers can focus on higher-value work.
  4. Design and build internal tools that make engineering more effective: productivity dashboards, AI-assisted issue triage, CI/CD optimizations, and tooling that reduces toil.
  5. Identify and remove organizational friction. Use data and direct observation to find where engineering is slowing down, recommend solutions to senior leadership, and build cross-team buy-in for changes.

Skills

Required

  • Experience building tools or workflows that improve how developers or AI agents work
  • Strong software and system design fundamentals
  • Genuine curiosity and hands-on experience with AI coding tools and agents
  • Excellent written and verbal communication; comfortable presenting to senior technical leadership
  • A track record of driving cross-team technical initiatives to completion

Nice to have

  • Experience with large-scale distributed systems or monolith decomposition
  • Prior work on developer experience or engineering productivity programs
  • Familiarity with code review tooling, static analysis, or automated PR pipelines

What the JD emphasized

  • AI-assisted coding domain
  • AI coding agents
  • internal systems they depend on need to be accessible via API
  • agent-ready
  • connections that let tools like Claude Code operate on them end-to-end
  • quality of AI-generated pull requests
  • automating the engineering work
  • context engineering standpoint
  • harness engineering standpoint
  • tools it can use
  • permissions it has
  • state it carries forward
  • tests it has to pass
  • logs you capture
  • retries, checkpoints, guardrails, and evals
  • AI harness tooling
  • AI-first coding workflows

Other signals

  • AI coding agents
  • internal systems accessible via API
  • build the connections that let tools like Claude Code operate on them end-to-end
  • improve the quality of AI-generated pull requests
  • automate engineering work