Research Scientist 5 — Content Representation Models (CRM)

Netflix
Los Gatos,California,United States of America2026-04-06onsite

About the job

We are looking for a Research Scientist specializing in embeddings and representation learning to investigate how we can enhance content understanding capability in Netflix's foundation models. How foundation models understand content is one of the most important open research questions for Netflix personalization. The person in this role will help shape that research direction and bring new ideas to the table.

Responsibilities

Drive applied research on enhancing content understanding capability in Netflix's foundation models

Conceptualize, design, implement, and validate new approaches to representation learning and content embeddings

Explore and apply state-of-the-art AI/ML techniques, including methods for improving how LLMs and foundation models represent and reason about content

Develop production-ready solutions and partner with application teams to ensure research translates into member-facing impact

Design and run rigorous offline experiments and evaluations to validate new approaches

Collaborate with cross-functional teams across content understanding, foundation models, and personalization applications

Contribute to the broader research community through publications at top venues

Qualifications

Minimum

Ph.D. in Computer Science or a related field with a strong publication record in embeddings, representation learning, or a closely related domain

3+ years of research experience with a track record of delivering quality results

Deep expertise in machine learning, including practical experience with LLMs and/or foundation models

Strong software engineering skills in Python (eg, PyTorch /)

Excellent communication and collaboration skills

Preferred

Experience in adopting LLM for Recsys. More specifically, building Semantic IDs and ground them in LLMs.

Experience in computer vision or multimodal AI

Industry experience in recommendation systems, search, personalization, or retrieval

Experience with LLM pre-training, fine-tuning, or distillation

Hands-on experience with distributed training

Publications in top ML conferences (NeurIPS, ICML, ICLR, KDD, RecSys)

Applied research experience in industrial settings