Sr. Data Scientist / Member Technical Staff II

Company Description

At Nielsen, we are passionate about our work to power a better media future for all people by providing powerful insights that drive client decisions and deliver extraordinary results. Our talented, global workforce is dedicated to capturing audience engagement with content — wherever and whenever it's consumed. Together, we are proudly rooted in our deep legacy as we stand at the forefront of the media revolution. When you join Nielsen, you will join a dynamic team committed to excellence, perseverance, and the ambition to make an impact together. We champion you, because when you succeed, we do too. We enable your best to power our future.

Job Description

We are seeking an experienced MTS 2 with a Data Science or Machine Learning background and 4+ years of experience to join our Ads AI team within Gracenote — a Nielsen Company. You will design and build intelligent data services that expose Gracenote's entertainment metadata to agentic AI systems for transparent, privacy-compliant CTV advertising targeting. Working at the intersection of data science and production engineering, you will ship ML-powered features as real-time services consumed by AI agents via the Model Context Protocol (MCP).

Responsibilities

• Design, build, and maintain the Ads MCP Server — exposing Gracenote's advertising metadata as structured tools and resources that AI agents can discover and invoke.

• Model and structure advertising data — define schemas, taxonomies, and retrieval strategies that make content metadata (genre, mood, brand safety, program attributes) useful for agentic ad targeting workflows.

• Develop ML-powered features — build models for content classification, segment recommendation, similarity matching, and contextual relevance scoring.

• Integrate with agentic AI workflows — work with LangGraph/LangChain-based orchestration to enable multi-step reasoning over advertising data.

• Optimize real-time data retrieval — ensure segment lookups and metadata queries meet latency requirements for programmatic bid-time decisions.

• Design prompts and structured output schemas that help LLMs reason effectively about advertising metadata.

• Analyze and improve data quality — identify gaps in metadata coverage, propose enrichment strategies, and validate model outputs against ground truth.

• Collaborate with data scientists, analysts, and engineers to implement and operationalize models and algorithms.

• Identify and drive process improvements: automating manual processes, optimizing data delivery, improving model evaluation pipelines.

• Create and maintain comprehensive technical documentation.

• Mentor junior engineers and provide technical leadership on ML and data-related decisions.

Qualifications

• 4+ years in software engineering with a data science or ML focus.

• Engineering degree in Computer Science, Data Science, ML, Statistics, or related field.

• Strong Python — async patterns, type hints, data pipelines.

• ML fundamentals — classification, NLP, embeddings, similarity search, recommendations, model evaluation.

• Data modeling — complex metadata schemas, taxonomy design, structured/semi-structured data at scale.

• Expert SQL and relational databases.

• LLM/GenAI — integrating large language models, structured output, and prompt engineering.

• REST API design serving data and ML outputs.

• Cloud platforms (AWS preferred) and their AI/ML services.

• Git and CI/CD practices.

• Data quality, governance, and validation principles.

Additional Information

Preferred skills :

• Model Context Protocol (MCP) or tool-use interfaces for AI agents.

• LangChain / LangGraph or similar agentic orchestration frameworks.

• AdTech domain — programmatic buying, OpenRTB, DSPs/SSPs, contextual targeting.

• Media metadata — content taxonomies (genre, mood, brand safety).

• Google Vertex AI / Gemini, AWS Bedrock, or similar cloud AI platforms.

• Vector databases / embedding retrieval.

• Big data (Spark, Flink) and pipeline tools (Airflow).

• Docker, Kubernetes.

• Real-time, low-latency data serving.

• MLOps — model versioning, monitoring, A/B testing, deployment pipelines.

• Infrastructure-as-code (Terraform, CloudFormation).

Please be aware that job-seekers may be at risk of targeting by scammers seeking personal data or money. Nielsen recruiters will only contact you through official job boards, LinkedIn, or email with a nielsen.com domain. Be cautious of any outreach claiming to be from Nielsen via other messaging platforms or personal email addresses. Always verify that email communications come from an @nielsen.com address. If you're unsure about the authenticity of a job offer or communication, please contact Nielsen directly through our official website or verified social media channels.