Tech Lead, Robustness and Performance

SambaNova is hiring a Robustness and Performance Tech Lead for SambaNova’s software organization. 

Opportunity

The third era of AI has arrived, powered by foundation models. Foundation models are achieving step-function increases in scale, versatility, and accuracy compared to legacy AI technologies, presenting an opportunity for organizations to fundamentally transform their business and operations. 

SambaNova DataScale is enabling organizations and enterprises to achieve the transformative promise of these new AI technologies with a fully integrated hardware-software system that delivers innovation across the full AI stack, creating the AI backbone for the next 10 years and beyond.

This role presents a unique opportunity to shape the future of AI and the value it can unlock across every aspect of an organization’s business and operations. It offers a full stack perspective of accelerating Foundational Large ML Models implementation through the deep learning software stack onto SambaNova’s hardware platforms.

This is a high-visibility position that will be at the core of improving the performance of Foundational Large ML Models on SambaNova’s platforms, working in conjunction with all the software teams. 

Responsibilities

• Study the software stack in its entirety and identify opportunities to make the software stack robust
• Understand large ML models and the aspects of their acceleration/ performance in accelerators
• Work with engineers across the software stack to devise and implement frameworks, methodologies, and tools that will help measure, instrument, and profile robustness across the software stack
• Work closely with the DevOps team to integrate robustness measures into integration pipelines
• Work closely with the hardware team to understand and adapt robustness into new software features
• Drive key programming model abstractions and AI primitives for robust support of current & future AI workloads
• Work with peers, domain experts, developers, customers, and work across the enterprise seeking optimal solutions
• Provide support for proposals in key areas aligned with core team competencies

Preferred Qualifications and Skills

• Experience with performance optimization and modeling of ML primitives on accelerators
• Experience with math and/or deep learning library development
• Experience with tools and scripting to boost developer productivity
• Fundamental understanding of key ML verticals: Vision, NLP, Recommendation
• An appreciation for the process and developing cross-disciplinary collaboration
• Excitement about high-performance systems engineering and performance debugging
• Experience with common compiler development practices and methodologies