Senior Software Engineer, Machine Learning, Discover Interest Modeling and Retrieval

Google•Mountain View, CA

2d•$174,000 - $252,000

About The Position

Google's software engineers develop technologies that help users connect, explore, and interact with information. This role is on Discover, Google's personalized interest feed, which serves hundreds of millions of daily active users. The role focuses on modeling each person's long-term interests, favorite creators, and in-the-moment intent. These models support large-scale personalization and downstream ranking. The role also contributes to Discover's retrieval and pre-ranking stack, which selects relevant candidates from a constantly refreshing corpus of articles, videos, creator posts, and AI-generated content. The stack operates in milliseconds at hundreds of millions of feeds per day. As a software engineer, you will work on a project important to Google's needs, with opportunities to take on new problems as the business evolves. The position involves building and launching machine-learning systems and contributing to a universally accessible and useful search experience.

Requirements

Bachelor’s degree or equivalent practical experience.
5 years of experience building and launching production machine-learning models, including 2 years of experience with recommendation, retrieval, search, or personalization systems.
1 year of experience designing, analyzing, and optimizing algorithms for high-performance systems, and developing and implementing machine-learning algorithms for practical applications.
Experience with large-scale data processing using Beam, Spark, or similar; Python, C++, and SQL; and JAX, TensorFlow, or PyTorch.

Nice To Haves

Master's degree or PhD in Computer Science or a related technical field.
5 years of experience with data structures and algorithms.
1 year of experience in a technical leadership role.
Experience developing accessible technologies.
A proven track record driving quality projects end to end, from design through A/B experimentation to launch.

Responsibilities

Write and test product or system development code.
Develop representation-learning models, including embedding or encoder models trained with contrastive, supervised, or self-supervised objectives, and set or sequence encoders such as attention pooling and Transformers.
Execute clustering or metric learning at scale, including hierarchical/agglomerative or density-based methods, and assess cluster quality.
Use large language models in machine-learning systems, such as LLM embeddings and autoraters for evaluation, and distill LLM judgments into classifiers.