Principal ML/AI Architect

Principal ML/AI Architect at Netflix

Netflix is one of the world's leading entertainment services, connecting over 300 million paid memberships in 190+ countries to a vast library of TV series, films, and games. Members enjoy the flexibility of playing, pausing, and resuming content anytime, anywhere, and can adjust their plans as needed.

The Opportunity

At Netflix, the core mission is to entertain the world by effectively connecting members with global stories. With over 270 million members, the product is designed to help users quickly discover great content. The AI for Member Systems (AIMS) organization is central to this experience, building and operating the advanced AI systems that power recommendations, personalization, search, discovery, and messaging. AIMS also drives the advancement of Netflix Foundation models.

The AIMS team leverages cutting-edge machine learning, Generative AI, and Large Language Models (LLMs) to deliver these experiences reliably and efficiently on a global scale. As the complexity and ambition of these systems grow, Netflix is significantly investing in a stronger ML infrastructure and architectural foundation. This includes shared capabilities, paved paths, and abstractions aimed at accelerating the development of AI-powered member experiences. We are actively seeking a Principal ML/AI Architect (L7) to provide deep technical leadership in this critical area.

This senior individual contributor role is focused on shaping the ML infrastructure, architectural patterns, and underlying systems that AIMS relies upon, ensuring they are cohesive, scalable, and of the highest quality.

Responsibilities

• Own the Architectural Vision for ML Infrastructure: Define and evolve the core architecture for AIMS' AI foundations, encompassing data and feature pipelines, training and evaluation workflows, online inference and serving, and shared services for recommendations, search, discovery, and messaging.
• Architect Paved Paths for AI Product Teams: Design the "paved road" for AIMS teams to efficiently build and deploy ML models and GenAI capabilities. This enables rapid iteration while maintaining consistent patterns in data access, training, evaluation, rollout, and monitoring.
• Sequence and Layer ML Capabilities Thoughtfully: Architect the integration of different layers of ML capability over time, such as embeddings and retrieval, ranking and personalization, LLM/GenAI components, and evaluation and safety. Make intentional decisions regarding the centralization versus decentralization of these capabilities.
• Create Reusable, Horizontal ML Components: Design new capabilities like representation services, evaluation frameworks, and LLM-powered features as reusable building blocks. These components should be extensible across multiple member experiences to prevent the formation of silos.
• Scope and De-risk New Architectural Directions: Explore novel technical directions through prototypes and proofs of concept, particularly in areas where optimal abstractions are not yet clear. Utilize hands-on work to validate approaches and inform long-term strategic decisions.
• Connect Dots Across AIMS Verticals: Understand how various AIMS pillars (Recommendations, Search & Discovery, Evidence/Evaluation, LLMs/Foundations, Messaging) utilize and extend ML infrastructure. Identify opportunities to consolidate, simplify, or generalize solutions.
• Shape Requirements with Platform and Infra Partners: Advocate for AIMS' specific ML infrastructure needs with broader Netflix platform and infrastructure teams. Translate these requirements into actionable capabilities and ensure architectural alignment with enterprise-wide platforms.
• Champion Technical Excellence and Best Practices: Elevate the standards for reliability, observability, performance, and cost-effectiveness across AIMS' ML systems. Guide the adoption of best practices (e.g., for evaluation, rollout, guardrails) and mentor senior engineers throughout the teams.

What We're Looking For

• Deep Experience with Large-Scale ML Infrastructure: Significant, hands-on experience in designing and building production ML systems, including robust data pipelines, efficient training/evaluation workflows, and online serving mechanisms for high-traffic, ML-driven products.
• Fluency with Modern ML and GenAI Patterns: Strong working knowledge of contemporary ML approaches such as recommendation/ranking systems and Generative AI/LLM applications. Ability to effectively design infrastructure that supports these advanced models.
• Hands-On Ability to Scope and Validate Architectures: Proven capability to independently build prototypes, explore new technologies, and translate ambiguous problems into concrete architectural proposals and reference implementations.
• Strength in Abstraction, Frameworks, and Reuse: Demonstrated ability to identify common patterns and design flexible, extensible, and easily adoptable frameworks and abstractions for engineers.
• Ability to Influence Technical Direction Across Teams: Comfortable collaborating with and influencing senior engineers and leaders across diverse teams. Skilled at building consensus and navigating complex trade-offs without direct formal authority.
• Comfort in a Fast-Moving, High-Context Environment: Thrives in an environment characterized by high autonomy and significant expectations. Capable of navigating ambiguity, making informed strategic bets, and iterating quickly to maximize AI's impact for Netflix members.

Key Skills/Competency

• ML Infrastructure
• Generative AI
• Large Language Models (LLMs)
• Machine Learning Systems
• Architectural Design
• Scalability
• Data Pipelines
• Recommendation Systems
• Personalization
• Technical Leadership