ML Compiler & System

Genesis AI
London
Workplace: HybridFull timeFunction: Software EngineeringSkills: ["Systems thinking","API design taste","Proactive ownership","Self-directed work","Test and benchmark discipline"]

Own Quadrants end to end—from the authoring API, IR, lowering, and optimization passes to numerical primitives and backend execution. Build a usability-first Python compiler that transforms NumPy/PyTorch-style code into high-performance GPU kernels. Drive system coherence as the project grows, benchmark real physics workloads in tandem with Genesis, and reduce regressions by maintaining correctness and performance across the stack.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
Genesis AI
Genesis AI
2 days ago

ML Compiler & System

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 5 hours agoStatus: Live

Job Summary

Own Quadrants end to end—from the authoring API, IR, lowering, and optimization passes to numerical primitives and backend execution. Build a usability-first Python compiler that transforms NumPy/PyTorch-style code into high-performance GPU kernels. Drive system coherence as the project grows, benchmark real physics workloads in tandem with Genesis, and reduce regressions by maintaining correctness and performance across the stack.
Location: London
Workplace: Hybrid
Employment Type: Full time
Job Function: Software Engineering

Key Responsibilities

  • •Own Quadrants end to end: API & authoring experience, IR, lowering, optimization passes, numerical primitives, and the roadmap.
  • •Design and maintain a stable lowering path from the computation graph constructs to multiple GPU backends.
  • •Implement and improve optimization for execution plans, including vectorization, memory/register usage, synchronization, and traffic elimination.
  • •Develop and maintain numerical primitives (e.g., Cholesky, eigendecomposition, matmul, reductions, solver kernels) with correctness and benchmarks.
  • •Track and hunt performance wins across Genesis and Quadrants, propose API changes, benchmark workloads, and catch regressions early.

Key Requirements

  • •Deep GPU understanding of what optimal execution looks like (access patterns, registers, shared memory, barriers, atomics, threading) and strong CPU intuition.
  • •Expertise translating authoring to kernels via compiler concepts (IR, lowering, optimization) while balancing maintainability.
  • •Strong idiomatic Python and API design taste for usability in numerical computing workflows.
  • •Ability to measure before optimizing and to target memory traffic and launch overhead prior to FLOPs.
  • •Proficiency building and validating with tests and benchmarks, along with regression tracking.
Skills:Systems thinkingAPI design tasteProactive ownershipSelf-directed workTest and benchmark discipline
Tech Stack:PythonNumPyPyTorchCUDAAMDGPUMetalVulkanPTXInline assemblyX86Arm64IRLoweringVectorizationRegistersShared memoryBarriersAtomicsThreadingCholesky

Company Brief

Genesis AI
Genesis AI is a global physical AI research lab and full‑stack robotics company building a universal robotics foundation model and horizontal platform to enable general‑purpose robots and scale automation of physical labor.
Industry: AI & Machine Learning
Company Size: Small (11 to 50 employees)
Growth: Early Stage Startup
Funding: Seed
Headquarters: Paris, France
Founded: 2024
WebsiteLinkedIn