Machine Learning, Recommendation Systems (Ads Infra and Ranking)

Meta•Bellevue, WA

1d•$154,003 - $217,000

About The Position

Meta's Ads organization builds systems that help businesses reach people likely to care about their offerings. The Ads Infra and Ranking organization owns the technical foundation for ad ranking, retrieval and serving, and the data, training, and inference infrastructure that supports them at Meta's scale. Its work includes deep learning architectures, sequence and representation learning, generative recommenders, distributed GPU training, model/system co-design, and large-scale low-latency serving, with a focus on compute efficiency, reliability, and privacy. As an ML Software Engineer in Ads Infra and Ranking, you will build production recommendation systems that improve ad relevance through retrieval and ranking. You will own model and software improvements from design and evaluation through rollout, partnering with data and serving engineers to deliver reliable personalization capabilities and measurable improvements.

Requirements

Bachelor's degree in Computer Science, Computer Engineering, a relevant technical field, or equivalent practical experience.
6+ years of programming experience in a relevant programming language, or 3+ years of experience plus a PhD.
At least 1 year of hands-on experience in one or more of machine learning, recommendation systems, large language models, or artificial intelligence.
Experience with scripting languages such as Python, JavaScript, or Hack.
Experience developing machine learning models at scale from inception to business impact.
Knowledge of developing and debugging in C/C++ and Java, or experience with scripting languages such as Python, Perl, PHP, and/or shell scripts.
Track record of setting technical direction for a team, driving consensus, and building successful cross-functional partnerships.
Proven experience designing, building, or deploying recommendation systems, such as collaborative filtering, content-based, or hybrid approaches, and personalization at scale.
Experience building and shipping high-quality work and achieving high reliability.
Experience improving quality through thoughtful code reviews, appropriate testing, proper rollout, monitoring, and proactive changes.

Nice To Haves

Direct experience with ads ranking or monetization, including auction dynamics, bidding, pacing, value modeling, or advertiser-facing optimization.
Experience with distributed training and large-scale low-latency model serving.
Master's degree or PhD in Computer Science or another ML-related field.
Experience with sequence modeling, multi-task/multi-objective optimization, and transfer learning for recommendations.
Experience with GPU scaling, model/system co-design, CUDA, Triton, or inference optimization.
Publications in top-tier conferences or journals, patents, or open-source contributions in recommendations or the LLM space.
Experience with privacy-preserving or aggregated measurement and delayed-conversion modeling.

Responsibilities

Own a significant area of the ads ranking stack end to end, including model architecture, features, training data, evaluation, and production rollout.
Lead architecture work in deep learning for recommendations, including sequence models, multi-task and multi-objective learning, transfer learning, and generative recommenders.
Define evaluation methodology and guardrails, including offline/online correlation, calibration, holdouts, and long-term and incrementality effects.
Drive value-model and auction-adjacent modeling work in partnership with auction, bidding, and pacing teams.
Lead GPU-scale training and serving efficiency work, including model/system co-design and inference optimization.
Set technical direction for a team, drive consensus, and lead successful cross-functional partnerships.
Raise engineering and modeling quality through code and design review, testing, staged rollouts, and monitoring.
Own reliability and responsible-AI practices for a major ranking surface.
Contribute to team and organizational growth through mentoring engineers, participating in recruiting and interviews, and identifying opportunities to improve team workflows.