Desktop Capture Engineer

About

Engine-level access with our proprietary multi-input capture software. That is how we describe what makes Origin Lab different, and building that software is one of the most demanding engineering challenges at the company.

We capture game worlds from the inside: synchronized video, keyboard and mouse input, camera telemetry, and depth from the engine's render pipeline. Our recorder runs in the field, on operators' machines, across a growing catalog of game titles and hardware configurations, with no IT support.

The role

You will own the Origin Lab recorder: a Tauri desktop application written in Rust and React/TypeScript. It is the tool our distributed network of human players uses to capture tasks, record synchronized sessions, and upload directly to our platform.

The recorder orchestrates video capture, communicates with our game engine plugins, captures input with millisecond precision, and handles large resumable uploads reliably. You will harden what exists and extend it as our capture capabilities grow. We need someone senior enough to own the technical direction of the recorder, not just execute tasks within it.

This role is based in Los Angeles, CA. We’re open to a hybrid work model for the right candidate.

What you'll work on

• Own the recorder codebase across the Rust backend and React/TypeScript frontend. The recorder is the product our operators live in, every decision you make shows up in their workflow.
• Keep video capture reliable in the field. The difference between a pipeline that works on your machine and one that works on a hundred machines you will never see is the whole problem. Ordering, timing, process lifecycle, hardware variance, that is your day.
• Own the capture pipeline end to end: coordinating multiple capture streams, streaming resumable uploads, and recovering cleanly when things go wrong mid-session.
• Integrate tightly with the plugin team so that telemetry, depth, video, and input stay synchronized across engines, titles, and hardware.
• Extend capture capabilities, new input sources, session annotations, operator workflows, and keep a field-deployed application reliable across every title and configuration we support.

What we're looking for

• 5 years of systems programming in Rust. async/await, error handling, and the borrow checker are comfortable territory.
• Built or maintained a Tauri or Electron desktop application in production, ideally on Windows. You have shipped installers, dealt with auto-update, and debugged things that only break on end-user machines.
• IPC and networked process communication, and you have actually debugged it in real desktop environments, not just read about it.
• Windows-specific APIs: input capture, process management, file system permissions, and the class of bugs that only appear with real hardware, real GPUs, and real anti-cheat software in the loop.
• Functional TypeScript/React for the UI layer, not a specialist, but capable of owning the frontend when it's your turn.
• A debugger-first mindset. You instrument and reproduce before you fix. You treat "it worked on my machine" as a starting point, not an answer.
• Shipped a product that non-technical users depend on. You understand the difference between code that works and a product that works.
• Comfort with agentic coding tools. We use Claude Code (and similar agentic systems) daily as part of how we ship, especially for the loop of pushing changes to a Windows test machine, iterating on capture behavior, and closing out compatibility issues fast. You should be fluent with these tools, opinionated about where they help and where they get in the way, and able to direct them effectively while still owning the code that lands.

Bonus: experience with video capture tooling, audio pipelines, or resumable upload infrastructure on AWS.

Origin Lab

Origin Lab acquires, creates, and delivers AI-enriched catalogs across 3D environments, video game capture, TV/Film, and animation. Rights-cleared, model-ready content, licensed, captured at the source, and delivered with the structure that world model training actually requires. We are working with AI researchers from Oxford, Google Research, and others to drive breakthroughs in Artificial World Intelligence.