Software Engineer, AI Dev Tools

Whatnot · Consumer · Los Angeles, CA · Engineering

Software Engineer focused on building AI development tools and infrastructure to enhance engineering productivity and enable autonomous agent capabilities within the company's codebase and processes. This role involves designing sandboxed environments, orchestration systems, and AI-aware infrastructure, acting as a primary interface between the Developer Tools team and the AI Tooling Working Group.

What you'd actually do

  1. Building and shipping AI tooling across the full spectrum—from near-term developer productivity wins to long-term autonomous agent capabilities—while helping shape what we prioritize next based on what you see working (and not working) on the ground
  2. Designing and building sandboxed, containerized environments where AI agents can operate autonomously—running code, executing tests, and iterating on solutions with appropriate guardrails
  3. Building multi-agent orchestration systems and workflows, including CI-fixing agents, SRE context agents, and other autonomous AI capabilities that reduce toil and accelerate delivery
  4. Enabling non-engineers (PMs, designers, operations) to leverage AI tools effectively through purpose-built workflows, onboarding experiences, and low-code/no-code interfaces
  5. Building and optimizing AI-aware developer infrastructure: [CLAUDE.md](http://CLAUDE.md) configurations, MCP servers, prompt engineering frameworks, and context management systems that make AI assistants more effective in our codebase

Skills

Required

  • 5+ years of software engineering experience
  • strong track record of building and shipping production systems
  • backend-leaning is fine, but you're comfortable getting around a frontend codebase too
  • broad knowledge of other AI and LLM products and services
  • genuine developer tooling enthusiast
  • built or improved internal tools
  • obsessed over local dev ergonomics
  • contributed to platform infrastructure
  • owned some of it in production
  • Deeply familiar with the current generation of AI coding tools—Claude Code (strongly preferred), Cursor, Codex, or similar
  • Strong product thinking
  • Clear communicator who adapts naturally to your audience

Nice to have

  • AI development tools, especially tools provided by the big 3 LLM companies (Anthropic, OpenAI, Google)

What the JD emphasized

  • primary contributor of our AI Tooling work end-to-end
  • build and own the AI tooling
  • not a machine learning or data science role
  • building the tooling, infrastructure, and workflows that allow AI coding agents and assistants to operate effectively
  • Critically, you must be a strong software engineer first
  • primary interface between the Developer Tools team and Whatnot’s AI Tooling Working Group
  • Deeply familiar with the current generation of AI coding tools—Claude Code (strongly preferred), Cursor, Codex, or similar—and use them to get real work done

Other signals

  • AI tooling
  • developer productivity
  • autonomous agent capabilities
  • multi-agent orchestration
  • AI-aware developer infrastructure