GPU Software Engineer

GPU Software Engineer – AI Infrastructure

Location: Bay Area (Onsite)

Level: Senior / Staff / Principal

We're partnered with a well-funded AI compute startup building a next-generation GPU platform for the demands of modern AI and data-centre computing.

This is an opportunity to help build a GPU software stack from the ground up, working across low-level systems software, GPU kernels, compilers and runtimes.

We're looking for engineers who love building things. People who are still writing code, solving difficult technical problems and actively embracing new AI-powered development tools to work faster and smarter.

Whether your background is in GPU compilers, runtime systems, kernel development or low-level GPU software, we'd be interested in speaking.

What You'll Be Doing

• Design and develop core components of a new GPU software platform.
• Build low-level GPU software, including kernels, runtimes, compiler infrastructure or execution systems.
• Work on challenging problems involving GPU compute, performance, scheduling and software/hardware interaction.
• Use AI coding tools and agents to accelerate software development, testing and iteration.
• Help establish modern, specification-driven and test-first development practices.
• Work closely with experienced GPU architects and engineers to influence the platform's technical direction.

What We're Looking For

• Strong hands-on software engineering experience, ideally in C++, CUDA or related low-level technologies.
• Experience in at least one of the following:
• GPU kernels and compute software
• GPU compiler development (LLVM, MLIR, Triton or similar)
• GPU runtimes and execution frameworks
• Low-level systems software for AI accelerators
• Data-centre GPU software infrastructure
• Genuine enthusiasm for AI-assisted software development, with practical experience incorporating AI tools into your workflow.
• Engineers who enjoy writing, testing, debugging and shipping code themselves.
• A builder's mindset and an interest in developing an entirely new computing platform.

Data-centre GPU experience is particularly attractive, although engineers from adjacent accelerator and systems software backgrounds are encouraged to apply.

Why Consider It?

• Build from the ground up: Help shape a new GPU platform rather than maintain an established codebase.
• High technical impact: Contribute directly to foundational architecture and engineering decisions.
• Experienced leadership: Work alongside engineers with deep experience developing industry-leading GPU technologies.
• Modern engineering culture: An environment embracing AI-assisted development and new approaches to software engineering.
• Early-stage opportunity: Join a well-funded business at a stage where individual contributions can make a substantial difference.

Apply through LinkedIn or contact [email protected] for a confidential conversation.

Recruitment managed by Morr0 — specialists in AI, semiconductors and advanced computing.