SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Attain is building a data-sharing ecosystem that empowers consumers to leverage their data in exchange for modern financial services, while enabling businesses to access real-time consumer insights for research and targeted advertising.
As VP of Data Science, you will own the scientific foundation of Attain's products and serve as the trusted authority when clients and partners stress-test the company's methodologies. Your core responsibility is ensuring that Attain's competitive advantage—defensible, rigorous measurement science—remains the standard across all offerings.
You will report to the SVP of Engineering and operate at the intersection of methodology, product development, and client trust. Your work spans three critical areas: (1) defining and maintaining standards for how Attain measures impact, proves causality, and validates modeled data; (2) owning the modeling behind core products including census balancing, attribution, audience targeting, and incrementality testing; and (3) translating advanced statistical methods into product decisions and business recommendations for commercial teams.
Key responsibilities include setting documentation standards for core calculations (attribution windows, touch-model methodology, ROAS conventions), representing Attain's methodology to engineering and commercial stakeholders both internally and externally, and developing a team of data scientists who combine methodological depth with product and client acumen. You will partner closely with Engineering and ML leadership on modeling infrastructure without directly owning that infrastructure, and manage engagements with external consultants who bring domain-specific expertise.
You are expected to have deep expertise in causal inference, incrementality, attribution, and media mix modeling, backed by strong statistical grounding in Bayesian methods and experimental design. Hands-on experience with large-scale consumer datasets (transaction, purchase, panel, or identity data) is essential, as is a career built in AdTech/MarTech with demonstrated ownership of methodology in front of both internal and external stakeholders. You should be a strong advisor who can clearly articulate every question, concern, and limitation in product or methodology choices, and you must be comfortable moving between technical ML/engineering audiences and commercial/executive ones with equal credibility. Proactive disclosure of model limitations is valued over defensive post-hoc justification.
