Data & AI Manager

Saatchi & Saatchi Health is hiring a Data & AI Manager for its Data & AI function within a pharmaceutical advertising agency. The role develops claims-based audience models and analyses for HCP and DTC campaigns, contributes to applied AI products and evaluation standards, and builds reusable analytics tooling while partnering with internal stakeholders.

Responsibilities

• Build the audiences. Develop HCP propensity, lookalike, and segmentation models on medical and pharmacy claims, plus the patient journey, line-of-therapy, switching, and adherence analyses underneath them. Deliver outputs in a form media and strategy partners can activate
• Write the SQL that makes it possible. Most of this work starts as a hard query against large, messy claims data. This is the core craft of the role
• Help build our AI products. Work on initiatives such as synthetic market-research persona panels, automated client reporting, and new-business audience generation, helping move them from working prototypes toward reliable production solutions
• Check the AI's work. Help define quality standards for AI-generated outputs, including audience narratives and physician panel simulations. Build evaluation sets and scoring frameworks that determine when outputs are trustworthy and when they are not
• Read campaign performance. Support the design of holdout, geo, and matched-market tests, and help interrogate the measurement our media partners deliver rather than taking it at face value
• Build dashboards and reusable tooling. Create scalable solutions so recurring analyses do not route through one person every time. Help establish standards and best practices as the team continues to grow
• Present to stakeholders. Partner with Strategy, Media, Account, and Creative teams and, over time, develop confidence presenting directly to clients

Preferred

• Healthcare claims experience with Symphony, IQVIA, or similar open-claims/APLD data, including fluency with ICD-10, NDC, NPI, and line-of-therapy logic. This is the single biggest accelerator into the role
• Pharmaceutical or healthcare marketing experience, including HCP and/or DTC marketing and an understanding of how MLR review and data-use restrictions shape what can be delivered
• Experience with consumer or digital data sources including Epsilon, Adobe Analytics, GA4, or similar datasets and platforms
• Experience with incrementality testing, attribution, or campaign measurement
• Experience building products or workflows using LLMs or routinely working AI-assisted as part of your development process
• Experience with Streamlit, dashboarding, programmatic deck generation, or other prototyping tools that help bring ideas to life quickly
• Healthcare claims data
• Symphony
• IQVIA
• Open-claims data
• APLD data
• ICD-10
• NDC
• NPI
• Line-of-therapy logic
• Pharmaceutical marketing
• Healthcare marketing
• HCP marketing
• DTC marketing
• MLR review
• Data-use restrictions
• Consumer data
• Digital data
• Epsilon
• Adobe Analytics
• Incrementality testing
• Attribution
• Campaign measurement
• LLM product development
• LLM workflow development
• AI-assisted development
• Streamlit
• Dashboarding
• Programmatic deck generation
• Prototyping tools

Benefits

• Medical coverage
• Dental coverage
• Vision coverage
• Disability coverage
• 401K
• Parental and family care leave
• Family forming assistance
• Tuition reimbursement
• Flexible time off
• Paid time off

Company

• Global marketing, communications, and digital transformation services.
• Paris, Ile-de-France, FRA
• https://www.publicisgroupe.com/
• Founded in 1926
• 10001+ employees

Qualifications

• 5-7 years in data science, marketing analytics, or a related analytical role
• Advanced SQL. You can write production-grade queries against a cloud warehouse (Amazon Redshift) over large, messy transactional data. This is the hard requirement
• Python for analysis and modeling, including pandas, scikit-learn, and some exposure to gradient boosting (XGBoost or similar). Version control as a habit
• A solid foundation in statistics and experimentation. You understand concepts such as holdouts, testing methodologies, and the distinction between correlation and causation
• Ability to explain complex analyses clearly to non-technical stakeholders and defend methodology when needed
• Curiosity about applied AI and enough hands-on exposure to have formed opinions about where LLM output succeeds and where it falls short
• Healthcare claims experience with Symphony, IQVIA, or similar open-claims/APLD data, including fluency with ICD-10, NDC, NPI, and line-of-therapy logic. This is the single biggest accelerator into the role
• Pharmaceutical or healthcare marketing experience, including HCP and/or DTC marketing and an understanding of how MLR review and data-use restrictions shape what can be delivered
• Experience with consumer or digital data sources including Epsilon, Adobe Analytics, GA4, or similar datasets and platforms
• Experience with incrementality testing, attribution, or campaign measurement
• Experience building products or workflows using LLMs or routinely working AI-assisted as part of your development process
• Experience with Streamlit, dashboarding, programmatic deck generation, or other prototyping tools that help bring ideas to life quickly
• 5-7 years in data science, marketing analytics, or a related analytical role.

Required skills

• Applied AI and Large Language Models, Campaign Measurement Attribution, Campaign Measurement Incrementality Testing, Consumer and Digital Analytics Adobe Analytics, Consumer and Digital Analytics Epsilon, Consumer and Digital Analytics GA4, Healthcare Claims Analytics IQVIA, Healthcare Claims Analytics Open-Claims/APLD Data, Healthcare Claims Analytics Symphony, Healthcare Codes and Data Standards ICD-10, Healthcare Codes and Data Standards NDC, Healthcare Codes and Data Standards NPI, Pharmaceutical and Healthcare Marketing DTC, Pharmaceutical and Healthcare Marketing HCP, Pharmaceutical and Healthcare Marketing MLR, Programmatic Deck Generation, Python XGBoost, Python pandas, Python scikit-learn, SQL (Amazon Redshift), Statistics and Experimentation, Streamlit, Version Control

Experience

• 5-7 years