Head of Data Science: Embeddings & Commercial AI

Jobtome; Linden (AL); On-site; USD 263,000 - 394,000; Full time

Job summary

Spectrum Intelligence Ventures seeks a Head of Data Science to lead the platform's transformer-based embedding model, develop predictive products, and guide a high-performing team. You’ll set evaluation standards, MLOps, and responsible AI practices while collaborating with product, sales, and external partners.

You will drive pilots and buyer-ready demonstrations, ensuring models meet performance and privacy requirements and translate methodology into measurable business value.

Qualifications

• Proven record leading data science for data, audience, identity, or measurement products in advertising/marketing.
• Hands-on predictive modeling on consumer-level data at scale (propensity, segmentation, lookalike).
• Strong grounding in measurement, incrementality, and causal inference.

Responsibilities

• Lead the platform's transformer-based embedding model and related products.
• Establish evaluation, MLOps, and responsible AI standards.
• Oversee data science team hiring, growth, and career development.
• Engage with customers/partners to translate methodology into commercial proof points.
• Drive pilots, evaluations, and revenue-impact initiatives with product and sales teams.

Skills

Data science leadership

Predictive modeling

Measurement & causality

Identity & data collaboration

End-to-end ML lifecycle

Python/SQL/Distributed processing

Deep learning & embeddings

Privacy-preserving analytics

Communicate with executives

Build high-performing teams

AI / foundation models curiosity

Education

Bachelor's in CS/Math/Stats/Economics

Job description

Spectrum Intelligence Ventures seeks a Head of Data Science to lead the platform's transformer-based embedding model, develop predictive products, and guide a high-performing team. You’ll set evaluation standards, MLOps, and responsible AI practices while collaborating with product, sales, and external partners.

You will drive pilots and buyer-ready demonstrations, ensuring models meet performance and privacy requirements and translate methodology into measurable business value.