Drimva

Simular

Software Engineer, CUA Control

Singapore +1 · On-site · Full-time · Posted Apr 23, 2026

Location
Singapore
Palo Alto

About the job

Build the cross-platform control layer that turns AI model intent into reliable computer actions and supports feedback and recovery.

What you'll do

  • Develop low-level input automation, screen capture, coordinate mapping, and input simulation.
  • Implement UI detection and cross-platform abstractions using accessibility APIs, DOM/a11y trees, and visual grounding.
  • Improve action reliability and recovery, and coordinate with model and planning teams on intent-to-execution interfaces.

What they ask for

  • Experience building OS-level input automation.
  • Understanding of accessibility frameworks on macOS, Windows, or Linux.
  • Experience addressing flaky selectors, timing issues, or resolution-dependent coordinates.
  • Careful attention to reliability and edge cases.

Nice to have

  • Experience with screen reader internals, RDP/VNC, game automation, LLM agent tool-use systems, or mobile device automation.
  • Experience with Playwright, Appium, PyAutoGUI, Hammerspoon, or similar tools.

Tools

CGEvent, SendInput, xdotool, AXUIElement, UI Automation, AT-SPI, Playwright, Appium

AI focus

Build computer-control systems that execute AI model intent and collaborate with the model and planning team; contribute to agent feedback and recovery loops.

Model-written summary · the original posting is authoritative

First seen Oct 27, 2026 · Last verified Nov 4, 2026 (UTC)