Quadrants: Compiler Lead

About This Opportunity

We are representing a confidential open-source infrastructure organization building next-generation compilation technology for high-performance numerical computing. Our client is developing a Python-to-machine-code translation layer that targets CPU and GPU backends, with a focus on physics simulation and performance-critical workloads.

This is a rare opportunity to own a compiler project end-to-end—from API design through IR, lowering, optimization passes, and numerical primitives. You will define the technical direction for an open-source compiler that aims to become the standard for performance-critical Python in research and production environments.

Responsibilities

• Own the compiler architecture, API surface, intermediate representation, and optimization pipeline across CPU (Arm64, x86) and GPU (CUDA, AMDGPU, Metal, Vulkan) backends
• Design and maintain the user-facing API: idiomatic Python decorators, clear constraints, sharp error messages, and escape hatches for advanced users who need low-level control
• Build and tune optimization passes: vectorization, memory layout, barrier and atomic placement, register allocation, shared-memory usage, dead-code elimination, and synchronization removal
• Implement and maintain numerical primitives (Cholesky, eigendecomposition, matmul, reductions, solver kernels) with correctness guarantees, benchmarks, and regression tracking
• Work closely with internal users to identify performance bottlenecks, API friction, and missing abstractions; propose changes that reduce complexity and improve adoption
• Track performance on real workloads over time, catch regressions early, and maintain the project as a product with clear benchmarks and user-facing documentation
• Grow into leading a small team (2–3 engineers) as the project scales

What We're Looking For

• Deep understanding of GPU architecture: memory access patterns, registers, shared memory, barriers, atomics, and cooperative threading; CPU optimization experience is a plus
• Strong compiler fundamentals: intermediate representations, lowering strategies, optimization pass design, and the tradeoffs between expressiveness and maintainability
• Fluency in Python and a sense for idiomatic API design—what feels natural to Python users and what creates friction
• System-level reasoning: you know where complexity should live, you measure before optimizing, and you think about memory traffic and launch overhead before FLOPs
• Proactive and self-directed: you gather context across teams, engage directly with users, and design in the open for audiences you may never meet
• Commitment to code quality: you write tests and benchmarks as you build, challenge assumptions in existing code, and treat understanding the system as real work
• Awareness of the competitive landscape and emerging techniques in compilation, GPU computing, and numerical libraries

How We Work With You

Candidates apply directly through Get A Job.ai. Our recruiting team will screen your application and conduct an initial interview. If there's a strong match, we will submit your profile to our client for consideration. Please do not attempt to contact the client directly—all communication will be coordinated through our talent team to ensure a smooth and confidential process.

Pay

Compensation details will be discussed during the screening process based on experience and location.

Equal Opportunity
Get A Job.ai is committed to inclusive hiring practices. We welcome candidates from all backgrounds and work to ensure fair and equitable access to opportunities.