Performance engineer

Performance engineer

Requirements

• Bachelor’s degree in Computer Science, Engineering, or a related field (Master’s preferred)
• 10+ years of experience in performance engineering, with a focus on large-scale distributed systems
• 2+ years of experience working with AI/ML technologies
• Proven experience in performance testing, profiling, and analysis of complex software systems
• Deep understanding of NLP architectures, training, and inference
• Experience with vector databases and search technologies
• Experience with cloud computing platforms (e.g., AWS, Azure, GCP) and containerization technologies (e.g., Docker, Kubernetes)
• Strong programming skills in Python
• Experience with performance analysis tools (e.g., profilers, debuggers, monitoring tools)

Responsibilities

• Define and implement performance engineering strategies for our Generative AI full stack, including services, application, LLMs, RAG pipelines, and related infrastructure
• Lead performance testing, profiling, and analysis efforts to identify and resolve performance bottlenecks
• Establish and maintain performance benchmarks and SLAs for critical AI services
• Provide technical leadership and mentorship to performance engineering team members
• Analyze and improve LLM inference performance, including latency, throughput, and resource utilization
• Develop and implement strategies for LLM capacity planning and scaling
• Collaborate with AI researchers to evaluate and improve LLM model architectures and training techniques for performance
• Optimize LLM inference through techniques such as quantization, distillation, and optimized kernel implementation
• Design and implement performance tests for RAG pipelines, including retrieval, ranking, and generation components
• Identify and optimize performance bottlenecks in RAG systems, such as database queries, vector search, and document processing
• Evaluate and optimize RAG system architectures for scalability and efficiency
• Tune vector databases for optimal recall and latency
• Collaborate with infrastructure teams to optimize hardware and software configurations for AI workloads
• Evaluate and recommend new technologies and tools for performance monitoring and analysis
• Develop and maintain performance dashboards and reports to track key metrics
• Optimize GPU utilization and memory management for LLM inference
• Work closely with AI researchers, software engineers, and product managers to ensure performance requirements are met
• Communicate performance findings and recommendations to stakeholders at all levels
• Stay up-to-date with the latest developments in Generative AI and performance engineering

Preferred Qualifications

No preferred qualifications provided.