Principal Machine Learning Scientist

We are UMG, the Universal Music Group. We are the world’s leading music company. In everything we do, we are committed to artistry, innovation and entrepreneurship. We own and operate a broad array of businesses engaged in recorded music, music publishing, merchandising, and audiovisual content in more than 60 countries. We identify and develop recording artists and songwriters, and we produce, distribute and promote the most critically acclaimed and commercially successful music to delight and entertain fans around the world.

How we LEAD:

Universal Music Group is building an Applied AI and Machine Learning function to help our teams discover and develop artists, connect with fans, make stronger commercial decisions, and protect the integrity of the music ecosystem.

We are looking for a Principal Machine Learning Scientist to turn that ambition into durable, high-impact products and decision systems. This is a senior individual-contributor role that combines hands-on technical work with technical leadership across cross-functional teams. It requires deep technical judgment, strong product sense, scientific and engineering rigor, and the ability to make progress in a complex, federated organization.

You will lead the design and delivery of AI-enabled workflows and machine learning solutions that improve decisions across UMG. Your work may span consumption and fan intelligence, catalog understanding, financial forecasting, rights, and ecosystem integrity.

You will work across business and technical teams to frame the right problem, choose the right method, and carry solutions through to measurable adoption and business impact. You will be equally comfortable developing a new model, challenging an existing approach, designing an AI evaluation framework, or helping a team turn a promising prototype into a reliable production capability.

How you’ll CREATE:

• •Turn ambiguous business opportunities into clearly defined machine learning, statistical, forecasting, optimization, generative AI, or non-AI problems.
• •Determine whether advanced modeling adds value or whether a simpler approach is more appropriate.
• •Lead initiatives from problem definition and data assessment through modeling, deployment, evaluation, monitoring, and sustained business adoption.
• •Build and improve models for areas such as fan intelligence, personalization, catalog understanding, forecasting, portfolio decisions, and fraud or content integrity.
• •Move generative and multimodal AI capabilities from prototypes into dependable, cost-effective workflows embedded in UMG products and operations.
• •Build decision systems that identify meaningful changes, explain supporting evidence, and help stakeholders determine appropriate action.
• •Define evaluation frameworks covering quality, latency, cost, robustness, privacy, provenance, copyright risk, and model safety.
• •Partner with ML engineers, data engineers, product managers, designers, domain experts, and label stakeholders to create systems people trust and use.
• •Establish reusable practices for data and feature access, experimentation, model serving, monitoring, governance, lineage, and documentation.
• •Communicate technical tradeoffs clearly to practitioners, executives, and business leaders, adapting the level of detail to the audience.
• •Influence cross-functional planning, connect capabilities across teams, and help prioritize work based on value, feasibility, risk, and learning potential.
• •Mentor practitioners and raise technical standards through reusable methods, documentation, and reviews.

What success looks like in your first 12 months

• •Lead one or more high-value AI or ML initiatives from ambiguous problem framing through production—or reach a well-evidenced decision not to build.
• •Establish a measurable evaluation and monitoring approach that stakeholders trust.
• •Improve the quality, reliability, cost efficiency, or adoption of an existing model or AI-enabled workflow.
• •Help a cross-functional team make a better business decision using evidence, not only a dashboard or prototype.
• •Raise the technical bar through reusable methods, documentation, reviews, and mentorship.
• •Build durable partnerships across central teams, labels, and business functions.

Bring your VIBE:

• •Advanced training in computer science, statistics, mathematics, economics, operations research, or another quantitative field, or equivalent applied research and industry experience.
• •Extensive experience developing and deploying machine learning or statistical systems, with sustained ownership of ambiguous, high-impact problems.
• •Deep production experience in at least two relevant areas, such as recommender systems, personalization, search and ranking, causal inference, time-series forecasting, optimization, or anomaly detection.
• •Strong experience designing, evaluating, and deploying modern AI systems, including LLM-based workflows, retrieval, grounding, tool use, structured generation, safety controls, or human review.
• •A strong record of moving models from research and experimentation into reliable, maintainable, and cost-efficient production environments.
• •Expert Python skills and fluency with modern scientific and ML tooling, such as PyTorch, TensorFlow, JAX, scikit-learn, or equivalent frameworks.
• •Experience with cloud-based data and ML platforms, such as GCP, AWS, or cloud-agnostic systems.
• •Strong understanding of data quality, feature design, experimentation, rigorous evaluation, and model monitoring.
• •A record of technical leadership, mentorship, and cross-functional collaboration without relying on formal authority.
• •Excellent written and verbal communication, including the ability to turn complex statistical or architectural tradeoffs into actionable guidance for executives and business leaders.

Especially valuable

Experience in music, media, entertainment, streaming marketplaces, advertising, or another industry where culture, consumer behavior, and business conditions change rapidly.

Experience working with complex non-tabular data, such as audio, text, images, or video.

Experience building applications for artist discovery, catalog understanding, fan marketing, rights management, fraud detection, or trust and safety.

Experience building decision-support products and automated workflows rather than standalone models or static dashboards.

Experience establishing model governance, data provenance, copyright safety protocols, or human-in-the-loop review systems.

Published research, open-source contributions, patents, or active collaboration with academic researchers and applied research labs.

Perks Playlist:

Join an entrepreneurial, global organization where authenticity, boldness, creativity, connection, drive, and insight aren’t just values—they’re how we work every day. Here are some of the ways we support you along the way (and just a few of the benefits we offer):

Comprehensive medical, dental, and vision coverage

Including 100% coverage for out-patient in-network mental health services

Fertility coverage for eligible medical plan participants

Wellbeing reimbursements for fitness classes, spa treatments, meal services, travel, and so much more (up to $720/year)

Student Loan Repayment Assistance and Tuition Reimbursement

401(k) with 100% immediate vesting on the first 5% of your contributions, plus an additional UMG contribution

A variety of ways to prioritize much-needed time away from work including:

Flexible Paid Time Off (PTO) for exempt employees

3-weeks PTO for non-exempt employees

2-weeks paid Winter Break

10 Company Holidays (including Juneteenth and Wellbeing Day)

Summer Fridays (between Memorial Day and Labor Day)

Generous paid parental leave for every type of parent

Check out our full overview of benefits on the Perks Playlist page of the career site.

Disclaimer: This job description only provides an overview of job responsibilities that are subje