Senior System Software Engineer, Agentic Retrieval

About the role

NVIDIA's technology is at the heart of the AI revolution, touching people across the planet by powering everything from self-driving cars, robotics, co-pilots and more. Join us at the forefront of technological advancement in intelligent assistants and information retrieval. NVIDIA is looking for a System Software Engineer - Agentic Retrieval to develop pipelines for indexing and querying multi-modal content. We are looking for someone with a passion for working with the world's most challenging problems in Generative AI, LLM, VLLM, and Agentic Retrieval spaces using our innovative hardware and software platforms. You will develop tools for building powerful, flexible, multi-modal retrievers and agents driven by Large Language Models(LLM) thereby improving the experience of millions of customers. If you're creative & passionate about solving real world conversational AI problems, come join us. This role is pivotal in accelerating containerized pipelines for high quality multi-modal datasets and providing best-in-class retrieval efficacy. The day-to-day focus is on developing efficient, scalable systems for deduplicating, filtering, and classifying training corpora for tailored models that enhance off-the-shelf capabilities. Fundamental to these efforts are iterative testing and improvement in system cost, speed, & accuracy through micro-optimization, prompt engineering, fine tuning, and applying new research. The ideal candidate believes in craftsmanship whereby they release early and often to obtain feedback while keeping the long-term vision alive! They are comfortable objectively evaluating the latest AI models and frameworks with an eye towards acceleration and capability enhancement.

Responsibilities

• Develop and optimize Rust-based data processing frameworks, ensuring efficient handling of large datasets on GPU-accelerated environments, vital for LLM training.
• Lead development and iterative optimization of components for Agentic Retrieval pipelines, ensuring they demonstrate GPU acceleration & the best performing models for improved TCO.
• Collaborate with teams of LLM & ML researchers in the development of full-stack, GPU-accelerated data preparation pipelines for multimodal models
• Implement benchmarking, profiling, and optimization of innovative algorithms in Python in various system architectures, specifically targeting LLM applications.
• Work closely with diverse teams to understand requirements, build & evaluate POCs, and develop roadmaps for production level tools and library features within the growing LLM ecosystem.