Senior Data Scientist

We seek a Senior Data Scientist to join our product-engineering data-science team and contribute to building agentic AI systems supporting NVIDIA's networking engineering workflows.

You will create and implement production-ready LLM-powered solutions that collaborate closely with networking and silicon engineering teams. These solutions will transform complex domain expertise into scalable, measurable automation combining data science and applied AI.

Responsibilities:

• Build and ship agentic systems featuring multi-agent coordination, long-horizon reasoning, and planning frameworks that drive real efficiency gains for networking and NW silicon engineers.
• Build full-scale domain-specific agents and retrieval pipelines (RAG) over diverse data sources, including product telemetry, development artifacts, and engineering knowledge bases.
• Partner with networking SMEs to translate rules of thumb into structured agent instructions, evaluation datasets, and feedback loops.
• Establish evaluation infrastructure using offline golden sets and live telemetry, enabling confident iteration, cost control, and safe rollout of improvements.
• Analyze agent failures and data gaps, then propose and prototype improvements to features, algorithms, and system development.
• Lay the technical groundwork for a reusable agent platform that scales beyond your immediate team.

Requirements:

• PhD/MSc in CS, CE, Data Science, Statistics, or a related field (or equivalent experience).
• 5+ years in applied ML/AI or large-scale systems, with 2+ years building agentic or LLM-powered applications in production environments.
• Proficiency in Python; hands-on experience with agentic frameworks such as LangChain, LangGraph, CrewAI, or similar.
• Hands-on fine-tuning experience across full-weight, LoRA, and PEFT methods, with a solid grasp of key concepts including scaling laws and advanced alignment techniques (RL, RLHF, RLAIF, KL-divergence regularization, and related methods).
• Hands-on NLP experience before the LLM era includes BERT, word embeddings, sequence models, and classical NLP pipelines.
• Solid ML foundations: supervised learning, feature engineering, model evaluation, and statistical modeling.
• Strong communication skills , able to align diverse collaborators, document assumptions clearly, and drive progress without formal authority.