Ground Truth Ops Specialist

About Moody’s

Moody’s brings together talented people to transform risk into opportunity. The company fosters an inclusive workplace where people can contribute openly, share ideas, and collaborate with colleagues and customers in a thoughtful, meaningful way. As a global leader in ratings and integrated risk assessment, Moody’s is also advancing AI so it can move from understanding complexity to taking action and helping clients respond to uncertainty with confidence.

Applicants are encouraged to apply even if they do not meet every requirement. Moody’s values strong relationships, curiosity, diverse viewpoints, practical action, and integrity.

Role overview

This position is responsible for managing the full lifecycle of ground-truth data collection programs. The goal is to convert machine learning needs into accurate, well-documented labeled datasets that support core computer vision products.

Skills and competencies

• Experience supporting or operating AI data labeling and annotation programs, including campaign coordination, quality assurance, or vendor oversight
• Practical exposure to professional labeling platforms such as Labelbox, Dataloop, Scale, or comparable tools
• Ability to analyze label spread, agreement metrics, and recurring error trends
• Working knowledge of Python for analysis and automation, including Jupyter notebooks
• Comfort using spreadsheets and performing basic analysis with SQL and/or Python
• Strong campaign and project coordination skills, including handling several active workflows at once
• Excellent attention to detail, especially in documentation, taxonomy design, and QA
• Clear written and spoken communication for working with technical teams and outside vendors
• Ability to give and receive feedback constructively to improve instructions and outputs
• Exposure to computer vision, geospatial data, or machine learning workflows is beneficial

Education

A master’s degree in a quantitative, analytical, or technical discipline is expected.

Responsibilities

• Work with ML and Data Science leaders to convert model needs into precise label taxonomies and task definitions
• Plan and run internal benchmark campaigns to validate taxonomies and cover edge cases
• Set up and oversee annotation projects, including uploading images, geometries, and metadata, and exporting completed data
• Manage relationships with external labeling vendors through training, examples, and ongoing feedback
• Assess vendor quality using gold-standard sets, confusion matrices, and reporting to judge scale readiness
• Start and track production labeling work while monitoring throughput, SLAs, and overall delivery progress
• Resolve worker questions, clarify taxonomy issues, and identify process or tooling problems early
• Track label quality and consensus, and trigger additional review rounds or manual QA when required
• Examine low-agreement cases and suggest taxonomy updates, clarifications, or removals
• Assemble final ground-truth datasets and ensure they meet quality standards and are clearly documented for ML use
• Keep campaign updates and documentation current so stakeholders have visibility into dataset readiness
• Help improve ground-truth workflows, tools, and operating practices over time

Team context

You will join Cape by Moody’s, a group building highly accurate property intelligence from aerial imagery and advanced machine learning. Ground-truth data is core to the team’s work, and the role collaborates closely with data scientists, ML engineers, and product partners across Europe and North America in a remote-friendly environment.

Equal opportunity and compliance

Moody’s is an equal opportunity employer. All qualified candidates are considered without regard to race, color, religion, sex, national origin, disability, veteran status, sexual orientation, gender expression, gender identity, or any other protected characteristic.

Candidates may be required to disclose securities holdings under Moody’s policy for securities trading and the requirements of the role. Employment is conditional on compliance with that policy, including any necessary remediation of holdings.