Skip to main content

Arcade AI

AI Engineer - Tools

San Francisco, CAfulltimemidAdded today

About this role

Arcade is hiring an AI Engineer to build and maintain tool integrations that enable AI agents to securely take real actions within enterprise systems. You'll design and implement API toolkits, ensure their reliability across different LLM models, and help establish patterns for the growing tool catalog powering production AI agents.

What you'll do

  • Design and build toolkits that wrap vendor APIs into clean, agent-callable interfaces
  • Write and maintain integration code against real third-party APIs
  • Create tests and evaluations proving tools work reliably across different models
  • Monitor and fix existing toolkit breakages as upstream APIs evolve
  • Tune tool descriptions and agent behavior to optimize LLM understanding and execution
  • Contribute to shared patterns and infrastructure that speed up future toolkit development

What they're looking for

  • Python and TypeScript
  • API design and integration
  • Software testing and quality assurance
  • LLM prompt engineering and tool descriptions
  • Agent behavior and evaluation
  • Distributed systems or production infrastructure experience
  • RESTful API knowledge
  • Documentation and technical communication
Apply with Autofill

Opens the application — the Jobs AI extension fills it for you. Set up autofill

Opens the official application on the employer’s site. No login required.

Arcade AI

Arcade AI builds infrastructure that enables AI agents to take real-world actions by connecting language models to tools, integrations, and workflows. The company is hiring Applied AI Engineers, Solutions Architects, and Distributed Systems Engineers to design agentic tools, enterprise architectures, and scalable runtime systems.

View all jobs at Arcade AI

Likely interview questions

  • Tell us about a time you integrated a complex third-party API—how did you ensure reliability in production?
  • How would you approach designing tool descriptions to help an LLM decide when and how to call a tool?