Machine Learning Analyst, Google Cloud Protection Analytics - Google
Sunnyvale, CA
About the Job
Minimum qualifications:
- Bachelor's degree or equivalent practical experience.
- 2 years of experience in data analysis, including identifying trends, generating summary statistics, and drawing insights from quantitative and qualitative data.
- 2 years of experience managing projects and defining project scope, goals, and deliverables.
Preferred qualifications:
- Master's degree in a quantitative discipline.
- 2 years of experience with one or more of the following languages (e.g,. SQL, R, Python or C++).
- 2 years of experience with machine learning systems.
- Excellent problem-solving skills with attention to detail in a fluid environment.
- Excellent written and verbal communication skills.
About the job
Google Cloud Platform (GCP), Google Protection Analytics (GPA) is Google Cloud’s team of abuse fighting and user trust experts working daily to make the internet a safer place. We partner with teams across Google Cloud to deliver solutions to stop abuse and provide safe and trusted experiences for our users.
As a member of the GCP Protection Analytics team, you will help deliver trusted experiences for Google Cloud users by building products that prevent abuse while ensuring Google Cloud customers are able to seamlessly grow on Cloud. You will leverage our user trust expertise, machine learning and problem-solving skills, and a cross-product perspective to build and deliver enterprise-ready, industry-leading solutions.
Responsibilities
- Analyze and solve problems using data and statistical methods.
- Identify and prevent fraud and abuse.
- Improve tools and automated systems through data analysis, technical expertise, and present to stakeholders.
- Perform data analysis to drive decision-making, such as monitoring the roll-out of new models, investigate the impact of machine learning model changes, and the root cause of unanticipated changes in key metrics.
- Evaluate new features and data sources to use in modeling, both to improve existing model performance and to adapt new potential model use cases.