Senior Remote Solution Architect: Kernel & ML Performance

Talanto; Kraków; Hybrid; PLN 320,000 - 600,000; Full time

Benefits offered by this job

Stable pay

Employee Stock Purchase Plan 15%

Health insurance

Multisport

Shopping vouchers

Referral bonuses up to $2,000

Relaxation zones and games at offices

Job summary

EPAM is recruiting a Solution Architect (Kernel Optimization & ML Performance) to shape high-performance ML solutions. You will drive architecture for performance-critical workloads, collaborating with ML researchers, compilers, and engineers to deliver scalable, production-grade systems.

You will guide cross-functional teams, influence framework choices, and lead benchmarking, autotuning, and profiling efforts, while upholding security and scalability standards.

Qualifications

• 12+ years of software engineering, incl. 5+ in architecture/leadership.
• Deep knowledge of ML frameworks (JAX, PyTorch, TF) and Core ML concepts.
• Hands-on kernel-level performance optimization for TPUs/GPUs.
• Strong C++/Python experience for HPC environments.
• Familiarity with ML compiler toolchains (MLIR/OpenXLA).
• Proven track record designing scalable production ML systems.
• Excellent stakeholder management and communication abilities.
• Experience with autotuning, benchmarking, and performance profiling.

Responsibilities

• Define and own architecture for performance-critical ML workloads with custom kernels on TPUs/GPUs.
• Create a strategic roadmap for kernel optimization and large-scale performance gains.
• Collaborate with client tech leadership, ML researchers, and engineers to capture requirements.
• Evaluate technologies, toolchains, and design patterns for compute-intensive ops.
• Lead benchmarking infrastructure, autotuning frameworks, and regression suites.
• Promote ML performance best practices across teams and frameworks.
• Provide technical leadership and mentorship to development teams.
• Ensure security, scalability, and maintainability in solution designs.

Skills

Architecture leadership

Kernel optimization

TPU/GPUs optimization

C++/Python HPC

ML frameworks knowledge

Communication skills

Open-source contributions

Education

Bachelor's or Master's in CS or equivalent

Tools

MLIR

OpenXLA

Job description

EPAM is recruiting a Solution Architect (Kernel Optimization & ML Performance) to shape high-performance ML solutions. You will drive architecture for performance-critical workloads, collaborating with ML researchers, compilers, and engineers to deliver scalable, production-grade systems.

You will guide cross-functional teams, influence framework choices, and lead benchmarking, autotuning, and profiling efforts, while upholding security and scalability standards.