About the job
Simular develops AI systems that interact directly with computers across different platforms. The Software Engineer - Control Layer Development focuses on the core layer that translates AI intent into real actions, managing everything from mouse and keyboard input to UI detection and error recovery on macOS, Windows, and Linux.
What you will do
- Design and maintain the low-level control stack for computer input, including mouse and keyboard injection, screen capture, and simulating user actions.
- Implement UI element detection using accessibility APIs (such as AXUIElement and UI Automation), DOM or accessibility trees, and visual grounding methods.
- Help build an abstraction layer so the agent operates smoothly across different operating systems and application types.
- Troubleshoot and resolve reliability issues related to UI changes, window occlusion, resolution scaling, and managing focus across applications.
- Develop feedback mechanisms so the agent can verify its actions and recover from unexpected events.
- Work closely with modeling and planning teams to improve how intent translates into execution.
Requirements
- Background in developing OS-level input automation tools (such as CGEvent, SendInput, xdotool).
- Strong knowledge of accessibility frameworks: AXUIElement for macOS, UI Automation for Windows, and AT-SPI for Linux.
- Experience handling unreliable element selectors, timing problems, and resolution-dependent coordinates.
- Careful approach to reliability and edge cases in software development.
- Familiarity with automation tools like Playwright, Appium, PyAutoGUI, Hammerspoon, or similar.
Bonus skills
- Understanding of screen reader internals.
- Experience with remote desktop protocols (RDP, VNC), game automation, LLM agent tool-use systems, or mobile device automation (iOS UIAutomation/XCTest, Android UIAutomator/Accessibility).
This position is based in Singapore.

