Senior Software Engineer (Game Engine)

Join Roblox as a Senior Software Engineer on the Geometry Scalability team. You'll design and ship 3D engine systems for avatars and environments, improve game engine performance across devices, and mentor other engineers. This role requires 8+ years of experience in high-performance systems, strong C++ proficiency, and the ability to work effectively across teams.

USD 243,290 - 295,250

Responsibilities

• Design and ship 3D engine systems for avatar and environment geometry, including mesh processing, level of detail, transcoding, streaming, runtime loading, and related performance systems.
• Profile and improve the performance of our game engine across a wide range of devices, with particular attention to mobile.
• Drive projects end-to-end, mentor other engineers, and help set a high bar for technical quality and code health on the team.
• Real-time Performance Mindset: Experience diagnosing tradeoffs among CPU/GPU time, memory, loading latency, visual quality, and frame-time stability on constrained platforms such as mobile or console
• Applied 3D Systems Experience: Demonstrated hands-on experience building at least one production system involving real-time 3D rendering, mesh or geometry processing, LOD, animation/skinning, spatial data, 3D asset pipelines, or GPU/client performance. Deep specialization in one area is not required, but meaningful hands-on experience is

Requirements

• Collaboration and Communication: Ability to work effectively across teams (Engine, Rendering, Avatar, Studio, and developer-facing teams) and explain complex technical topics clearly to different audiences
• Proven Expertise: 8+ years of professional experience building and shipping high-performance systems in domains such as game engines, real-time simulation, robotics, AR/VR, or similar
• Strong C++ Experience: Deep proficiency in C++, with a focus on writing clean, efficient, and maintainable code in large codebases
• Ownership in Ambiguity: A track record of driving work end-to-end when the problem wasn't fully specified: defining the approach yourself, seeing it through implementation, rollout, and production consequences; gathering more information when requirements didn't make sense