Build high-performance machine learning infrastructure spanning compilers, runtimes, optimizers, and heterogeneous hardware, accelerating research iteration and production workloads through deep systems-level engineering.

What You'll Do:

1. Build optimizers, code generators, runtime libraries, and computation orchestration systems.
2. Improve the speed and efficiency of machine learning research and production workloads.
3. Design systems that execute computation across heterogeneous hardware.
4. Work across the entire compute stack, from high-level orchestration to hardware-level performance.
5. Analyze compute graphs and identify opportunities to optimize critical execution paths.
6. Investigate compiler behavior and understand how software translates into hardware execution.
7. Develop solutions using technologies such as C++, Rust, Python, MLIR, and CUDA.
8. Collaborate with stakeholders to understand requirements and turn them into effective technical solutions.

Your responsibilities will include:

1. Designing and developing machine learning framework infrastructure.
2. Optimizing runtimes and compute pipelines for maximum hardware performance.
3. Building tooling that accelerates research iteration and experimentation.
4. Working with compilers, code generation, runtime systems, and orchestration layers.
5. Profiling and analyzing hardware behavior to identify performance bottlenecks.
6. Improving hot paths and extracting maximum performance from available hardware.
7. Moving fluidly between high-level system architecture and low-level implementation details.
8. Learning unfamiliar technologies and areas of the stack independently.
9. Evaluating technical approaches and selecting the right language or tool for each problem.
10. Working closely with stakeholders across the organization to identify underlying needs and deliver practical solutions.

Why Join Us:

1. Work across theentire machine learning compute stack , from orchestration to hardware execution.
2. Solve challenging problems involving compilers, runtimes, heterogeneous computing, and performance optimization.
3. Balance rapid research iteration with production-grade performance.
4. Work with modern technologies including C++, Rust, Python, MLIR, and CUDA.
5. Explore problems at both architectural and extremely low-level hardware layers.
6. Have the opportunity to optimize systems where every increment of performance matters.

About You:

1. Comfortable moving between multiple programming languages and technologies.
2. Pragmatic about technology and focused on choosing the right tool for the problem.
3. Curious about how programming languages compile and execute on hardware.
4. Comfortable investigating unfamiliar systems and quickly developing a working understanding.
5. Interested in both high-level architecture and low-level performance engineering.
6. Strong analytical and problem-solving abilities.
7. Comfortable working with complex technical systems and performance-critical code.
8. Effective at collaborating with stakeholders across different areas of the organization.
9. Able to identify the underlying requirement behind a stakeholder's initial request.
10. Excited by challenging problems where the solution is not always obvious.

#MachineLearning #MLFrameworks #SoftwareEngineering #PerformanceEngineering #CPlusPlus #Rust #Python #CUDA #MLIR #Compilers #CodeGeneration #DistributedSystems #HeterogeneousComputing #HighPerformanceComputing #MachineLearningInfrastructure