Senior Inference Systems Performance Architect

SambaNova; San Jose (CA); On-site; USD 245,000 - 325,000; Full time

Benefits offered by this job

Health insurance

Gympass+

One Medical

Job summary

SambaNova Systems in San Jose, CA, seeks an Architect for the Inference Systems Performance team to own end-to-end performance for large-scale LLM inference, from tokenization to decode, across hardware and software. You will define strategy and drive reproducible workloads to optimize efficiency.

You will model, simulate, and optimize configurations to meet customer SLOs, mentor engineers, and represent our performance story to customers and partners.

Qualifications

• 12+ years of experience in performance engineering and leadership.
• End-to-end performance analysis of distributed systems.
• Realistic workload generation and performance modeling experience.
• Ability to enter unfamiliar domains and apply core methods.
• Lead cross-functional efforts and influence direction.
• Represent the organization credibly to customers and partners.
• Deliver high-complexity work with significant product impact.

Responsibilities

• Define and drive technical strategy for inference-systems performance.
• Build workload-capture and benchmarking capabilities.
• Own performance-modeling and simulation to guide capacity planning.
• Develop end-to-end profiling tools for distributed inference pipelines.
• Serve as senior technical voice across teams and stakeholders.
• Represent SambaNova's performance story to customers and partners.
• Mentor senior engineers and scale capability across the org.
• Resolve ambiguous challenges spanning organizational boundaries.

Skills

Performance engineering

Distributed systems

Workload generation

Performance modeling

Technical leadership

Cross-functional collaboration

Customer/partner interaction

Tools

Workload simulation tools

Profiling tools

Job description

SambaNova Systems in San Jose, CA, seeks an Architect for the Inference Systems Performance team to own end-to-end performance for large-scale LLM inference, from tokenization to decode, across hardware and software. You will define strategy and drive reproducible workloads to optimize efficiency.

You will model, simulate, and optimize configurations to meet customer SLOs, mentor engineers, and represent our performance story to customers and partners.