Lead ML Architect for High-Scale Ads & AI

Google; Pittsburgh (Allegheny County); On-site; USD 207,000 - 300,000; Full time

Job summary

Google is seeking a senior software engineer focused on large-scale ML systems for ads. The role involves building low-latency pCTR architectures, leveraging TPU capabilities, and collaborating with DeepMind and Google Research to translate academic insights into production.

You will shape loss functions, calibration methods, and agentic ML workflows to accelerate model discovery and deployment across Search and Shopping Ads.

Qualifications

• Bachelor's degree or equivalent practical experience.
• 8 years of software development, incl. 5 years with large-scale ML/DL/NN/recommendation systems.
• Experience designing large-scale production DL/NN architectures under latency constraints.
• Experience leading cross-functional technical projects and mentoring engineers.

Responsibilities

• Lead technical architecture, delivery, and cross-team strategy for Search and Shopping Ads pCTR models.
• Design, prototype, and scale high-capacity pCTR architectures for TPU performance and low latency.
• Develop modeling solutions to capture deep user history and attention signals in AI Search experiences.
• Engineer loss functions and calibration methods to improve top-line metrics and auctions.
• Build agentic ML workflows to accelerate model architecture and feature space discovery.

Skills

Large-scale ML experience

Production DL/NN architectures

Leadership & mentoring

Education

Bachelor's degree or equivalent practical experience

PhD (preferred)

Tools

Ads prediction systems (AdBrain/Admixer)

TPU optimization

Job description

Google is seeking a senior software engineer focused on large-scale ML systems for ads. The role involves building low-latency pCTR architectures, leveraging TPU capabilities, and collaborating with DeepMind and Google Research to translate academic insights into production.

You will shape loss functions, calibration methods, and agentic ML workflows to accelerate model discovery and deployment across Search and Shopping Ads.