Head of Data Science: Embeddings & Audience Intelligence

Jobtome; Town of Hempstead (NY); On-site; USD 263,000 - 394,000; Full time

Job summary

Spectrum is seeking a Head of Data Science to lead our transformer-based embedding platform, guiding predictive models across audiences, identity, and measurement domains. The role requires building a high‑performing team, setting evidence standards for buyers, and driving pilots to contracts with cross‑functional partners in product and sales.

You will own the embedding model roadmap, establish MLOps and responsible AI standards, and work with external partners on architecture, training, and

Qualifications

• Proven record leading data science for data, audience, identity, or measurement products in advertising/marketing.
• Deep hands-on expertise in predictive modeling on consumer-level data at scale, including propensity, segmentation, and lookalike modeling.
• Strong grounding in measurement, causal inference, and test design (incrementality, controls, holdout).
• Knowledge of identity graphs, clean rooms, and third-party data usage in modeling.
• Experience with the full data science lifecycle from feature engineering to deployment at scale.

Responsibilities

• Build and run predictive models behind the platform's audience and data products.
• Own segment definitions, coverage, refresh cadence, and measured accuracy for audiences.
• Lead the platform's Feature Store as a commercial product with documented features.
• Lead the outcome anchoring program training models against licensed third-party outcome data.
• Lead modeling for custom and partner-specific audiences, including seed-based lookalike expansion.
• Establish repeatable, clean-room workflows with transparent match rates and performance against buyer baselines.
• Design and defend the platform's measurement methodology with controlled experiments and validated outcomes.
• Define metrics that power insights subscriptions and aggregate demand indices.
• Lead the data science behind identity and viewing products, including embeddings and transformer models.

Skills

Predictive modeling

Measurement and causal inference

Python

SQL

Distributed data processing

Communication to executives

Education

Bachelor's Degree in Computer Science, Statistics, Mathematics, Economics, or a related quantitative field

Tools

Python

SQL

Spark

Snowflake

Databricks

AWS

Job description

Spectrum is seeking a Head of Data Science to lead our transformer-based embedding platform, guiding predictive models across audiences, identity, and measurement domains. The role requires building a high‑performing team, setting evidence standards for buyers, and driving pilots to contracts with cross‑functional partners in product and sales.

You will own the embedding model roadmap, establish MLOps and responsible AI standards, and work with external partners on architecture, training, and