Google Ads is looking to improve the next generation of its advertising system by developing RoboRating capabilities using Large Language Models (LLMs). This involves creating reliable and reproducible ratings for various real-world tasks, personalizing these ratings to target users, and designing simulation systems that accurately model user behavior and journeys.
Requirements
- Experience with distributed Machine Learning and Machine Learning infrastructure, distributed systems and with Machine Learning algorithms at scale.
- Experience with applied computer vision or Machine Learning research and development, and supporting infrastructure.
- Experience with Large Language Models (LLMs).
- Experience in Machine Learning, or research achievements in the field of AI (e.g., product ownership, publications).
Responsibilities
- Understand current academic literature on LLM models' capabilities, and research on autorating (through LLMs, and otherwise), user simulation, and evaluation.
- Understand the current system, and the needs for rating and simulation tasks in the Google Ads context.
- Build and deliver functioning systems to Google Ads, with the goal to improve the next generation of advertising system, for the user, the advertiser, while keeping in mind Google's serving constraints and systems.
- Creating RoboRating capabilities with existing Large Language Models (LLMs) that produce reliable, reproducible ratings for real-world tasks.
- Designing simulation systems that match or beat on real world user journeys.
- Rating the journey produced against a slate of evaluation criteria, like helpfulness to the user, suitability to the task, keeping the user persona in mind.
Other
- Bachelor's degree in Computer Science or equivalent practical experience.
- PhD degree in Computer Science, Artificial Intelligence, Machine Learning, or a related technical field, or equivalent practical experience.
- The US base salary range for this full-time position is $166,000-$244,000 + bonus + equity + benefits.
