Associate Director, Commercial AI and Advanced Analytics- Oncology
for each heading. Use - for lists. Use
for prose. Do not repeat job title. Do not include Job ID etc. Remove boilerplate: cookie/navigation/legal/EEO etc. Also remove "How We Work", "Supporting People With Disabilities", "Candidate Rights", "Data Protection", etc. Also remove "Compensation Overview" heading? We'll keep Pay section with the range. Benefits section with description.
We need to preserve all substantive information. So include summary text, responsibilities bullet list, qualifications bullet list, benefits paragraph, pay paragraph.
We need to ensure no filler like "commensurate with experience". Not present.
We need to output only HTML fragment.
Let's extract summary: After "Summary" line: "The Associate Director, Commercial AI and Advanced Analytics is responsible for delivering advanced commercial measurement and decision intelligence across BMS brands and therapeutic areas. This role owns the design, execution, and operationalization of advanced Marketing Mix (MMx), causal, and optimization analytics that inform investment decisions, omnichannel strategy, and performance optimization. This position is highly hands‑on and execution‑focused, combining strong technical depth in marketing analytics with the ability to translate complex modeling outputs into clear, actionable recommendations for Marketing, Brand, Omnichannel, Field, and Finance partners. The Associate Director operates within an enterprise analytics framework, aligning with TA Analytics Leads and centralized platforms while ensuring high‑quality, on‑time delivery of insights that directly impact brand strategy and ROI."
That's the summary.
Responsibilities: There are headings: "Deliver Advanced Marketing & Commercial Measurement", "Build & Operate Advanced MMx Capabilities", "Agentic & Always‑On Insights Delivery", "Governance, Compliance & Quality", "Stakeholder Partnership". Under each there are bullet-like sentences separated by periods. We need to convert to list items. Probably each sentence is a responsibility. We'll create a
with - for each distinct responsibility. Could group under subheadings? But rule: use
for section headings only. No subheadings inside. So we can just have one Responsibilities section with a list of all responsibility sentences.
Let's list them:
From "Deliver Advanced Marketing & Commercial Measurement":
- Own end‑to‑end development and execution of advanced Marketing Mix (MMx) and Key Driver Analyses (KDA) to support brand planning, budget optimization, and performance assessment.
- Apply advanced modeling techniques to quantify incremental impact across channels (sales force, digital, TV, congresses, patient programs).
- Translate modeling outputs into clear business implications for brand strategy, media mix, and investment allocation.
- Support annual brand planning, in‑year optimization, and scenario analysis using MMx outputs.
From "Build & Operate Advanced MMx Capabilities":
- Design and run advanced MMx models using Bayesian, Huber, etc. approaches, including:
- Adstock and carryover modeling
- Saturation and diminishing returns curves
- Distributed lag and non-linear response functions
- Cross‑channel interaction effects
- Develop scalable and repeatable MMx pipelines aligned with enterprise standards and automated platforms.
- Partner with centralized teams to integrate MMx into self‑service and agent‑enabled analytics solutions.
- Ensure models are explainable, auditable, and fit for use in a regulated pharmaceutical environment.
From "Agentic & Always‑On Insights Delivery":
- Contribute to the design and execution of agentic and always‑on insights capabilities that provide continuous decision support to Brand, Omnichannel, and Field teams.
- Operationalize MMx, KDA, and causal models into always‑on analytics pipelines with defined refresh cadence, data quality checks, and monitoring.
- Structure model outputs (drivers, sensitivities, scenarios, constraints) to be consumable by AI‑enabled decision tools and agent‑based workflows.
- Support “what changed / why” and forward‑looking recommendation logic using causal and driver‑based methods.
- Ensure transparency, lineage, and explainability from data → model → insight → recommendation, with appropriate human‑in‑the‑loop review.
- Define and maintain the Key Business Questions (KBQ) framework and business context layer, bridge business intent with the underlying technical semantic layer.
- Partner with centralized platform, engineering, and Hyderabad teams to define inputs, features, and guardrails for agent‑enabled insights.
- Design insights to be role‑based, actionable, and embedded into brand planning, optimization, and omnichannel execution workflows.
- Drive adoption of agentic AI tools through role-based training, and structured feedback; serve as the primary business point of contact and channel user feedback into roadmap iteration.
From "Governance, Compliance & Quality":
- Adhere to data privacy, AI governance, and model validation standards (HIPAA, GDPR/CCPA, internal AI controls).
- Maintain clear documentation of assumptions, methods, and limitations.
- Participate in model reviews and audits with Legal, Privacy, and Governance stakeholders.
From "Stakeholder Partnership":
- Act as a trusted analytics partner to Brand, Marketing, Omnichannel, Field, and Finance teams.
- Present insights in concise, business‑first narratives suitable for brand leadership and planning forums.
- Support TA Analytics Leads by sharing best practices, templates, and modeling approaches.
We need to keep sub-bullets? The rule: use - for any list of duties/requirements/benefits. Could nest lists? Probably okay but simpler to flatten. However sub-items like under "Design and run advanced MMx models using Bayesian, Huber, etc. approaches, including:" then list of four items. We could keep as separate
- items, maybe include the leading phrase as part of li. Eg: "Design and run advanced MMx models using Bayesian, Huber, etc. approaches, including: Adstock and carryover modeling, Saturation and diminishing returns curves, Distributed lag and non-linear response functions, Cross‑channel interaction effects." Could combine. But better to keep each as separate li for clarity.
We'll produce a flat list; each bullet is a responsibility.
Qualifications: list items from qualifications section.
Extract qualifications bullet points:
- Advanced degree in Statistics, Economics, Data Science, Operations Research, or a related quantitative field.
- Minimum 5 years of experience in pharmaceutical commercial analytics, with direct ownership of MMx / MMM and causal analysis, or advanced ML models.
- Proven experience supporting brand strategy, budget planning, and performance measurement.
- Hands‑on expertise in advanced Marketing Mix Modeling, including:
- Bayesian and hierarchical modeling frameworks
- Adstock, lag, and saturation modeling
- Non‑linear optimization and scenario simulation
- Strong background in causal inference and incrementality measurement, such as:
- Geo‑experiments and matched‑market tests
- Synthetic control methods
- Uplift / incremental response modeling
- Proficiency in Python and/or R for statistical modeling, feature engineering, and automation.
- Solid SQL skills and experience working with large, multi‑source pharma datasets.
- Experience with cloud analytics platforms (e.g., Databricks, Snowflake) and production data pipelines.
- Familiarity with MLOps, version control, and model lifecycle management.
- Ability to translate analytical requirements into agent specifications, KBQ frameworks, and context layer definitions
- Experience leading or executing UAT for analytics or AI-enabled platforms, including test plan design, execution, and stakeholder coordination
- Ability to explain complex models and uncertainty to non‑technical stakeholders.
- Comfortable working with ambiguity, able to make sound analytical and business judgments when problem definitions, data, or requirements are not fully defined
- Product-minded analytical thinker, able to bridge the gap between analytical rigor and user experience design
- Domain & Platform Knowledge: Strong understanding of pharma commercial data (claims, APLD, specialty pharmacy, promotional, and digital engagement data).
- Experience integrating analytics with BI tools and decision workflows.
- Exposure to AI‑enabled or automated analytics platforms is a plus (agentic systems not required but beneficial).
- Demonstrated ability to drive end-user adoption of new analytics or AI tools through training, enablement, and change management
We'll flatten similarly.
Benefits: paragraph from Benefits section.
We need to include the benefits description: "Subject to the terms and conditions of the applicable plans then in effect, you may be eligible to participate in our comprehensive benefit plans – including wellbeing support, retirement and financial protection benefits, and insurance offerings (medical, dental, vision, life and disability). U.S.-based exempt employees are eligible for Flexible Time Off (FTO), which provides paid time off without a set accrual limit, subject to manager approval, along with 11 paid company holidays each year. Non-exempt employees, RayzeBio employees, and employees located in Puerto Rico receive 160 hours of paid vacation annually for new hires (subject to manager approval), 11 paid company holidays, and 3 optional holidays. Depending on eligibility, employees may also have access to additional time-off benefits, including paid sick leave, up to two paid volunteer days per year, summer hours flexibility, and leaves of absence for medical, personal, parental, caregiver, bereavement, or military needs. Eligible employees also enjoy an annual Global Shutdown between Christmas Day and New Year's Day. U.S.-based job seekers can explore full benefit offerings at https://careers.bms.com/benefits"
We'll keep as
.
Pay: from Compensation Overview: "Princeton - NJ - US: $167,540 - $203,013 The starting pay range(s) listed above is for full-time employees (FTE). You may also be eligible for additional discretionary incentive cash and stock opportunities. We determine starting pay thoughtfully – carefully considering the nature of the role, required skills, work location, schedule and the knowledge and experience you bring. Final compensation is guided by pay equity principles and applicable employment laws. Compensation programs are reviewed on an ongoing basis and may be adjusted over time to reflect evolving market factors, and individual, team or Company performance."
We'll put that in
under Pay heading.
We should not include schedule section because none.
Now produce HTML:
About the role
...summary...
Responsibilities
- ... each responsibility ...
...
Qualifications
- ... each qualification ...
...
Benefits
... benefits paragraph ...
Pay
... pay paragraph ...
Make sure no extra whitespace? It's fine.
Now craft the list items.
Responsibilities list items (I'll write them concisely but preserve meaning):
1. Own end‑to‑end development and execution of advanced Marketing Mix (MMx) and Key Driver Analyses (KDA) to support brand planning, budget optimization, and performance assessment.
2. Apply advanced modeling techniques to quantify incremental impact across channels (sales force, digital, TV, congresses, patient programs).
3. Translate modeling outputs into clear business implications for brand strategy, media mix, and investment allocation.
4. Support annual brand planning, in‑year optimization, and scenario analysis using MMx outputs.
5. Design and run advanced MMx models using Bayesian, Huber, etc. approaches, including: Adstock and carryover modeling, Saturation and diminishing returns curves, Distributed lag and non-linear response functions, Cross‑channel interaction effects.
6. Develop scalable and repeatable MMx pipelines aligned with enterprise standards and automated platforms.
7. Partner with centralized teams to integrate MMx into self‑service and agent‑enabled analytics solutions.
8. Ensure models are explainable, auditable, and fit for use in a regulated pharmaceutical environment.
9. Contribute
for prose. Do not repeat job title. Do not include Job ID etc. Remove boilerplate: cookie/navigation/legal/EEO etc. Also remove "How We Work", "Supporting People With Disabilities", "Candidate Rights", "Data Protection", etc. Also remove "Compensation Overview" heading? We'll keep Pay section with the range. Benefits section with description. We need to preserve all substantive information. So include summary text, responsibilities bullet list, qualifications bullet list, benefits paragraph, pay paragraph. We need to ensure no filler like "commensurate with experience". Not present. We need to output only HTML fragment. Let's extract summary: After "Summary" line: "The Associate Director, Commercial AI and Advanced Analytics is responsible for delivering advanced commercial measurement and decision intelligence across BMS brands and therapeutic areas. This role owns the design, execution, and operationalization of advanced Marketing Mix (MMx), causal, and optimization analytics that inform investment decisions, omnichannel strategy, and performance optimization. This position is highly hands‑on and execution‑focused, combining strong technical depth in marketing analytics with the ability to translate complex modeling outputs into clear, actionable recommendations for Marketing, Brand, Omnichannel, Field, and Finance partners. The Associate Director operates within an enterprise analytics framework, aligning with TA Analytics Leads and centralized platforms while ensuring high‑quality, on‑time delivery of insights that directly impact brand strategy and ROI." That's the summary. Responsibilities: There are headings: "Deliver Advanced Marketing & Commercial Measurement", "Build & Operate Advanced MMx Capabilities", "Agentic & Always‑On Insights Delivery", "Governance, Compliance & Quality", "Stakeholder Partnership". Under each there are bullet-like sentences separated by periods. We need to convert to list items. Probably each sentence is a responsibility. We'll create a
- with
- for each distinct responsibility. Could group under subheadings? But rule: use
for section headings only. No subheadings inside. So we can just have one Responsibilities section with a list of all responsibility sentences. Let's list them: From "Deliver Advanced Marketing & Commercial Measurement": - Own end‑to‑end development and execution of advanced Marketing Mix (MMx) and Key Driver Analyses (KDA) to support brand planning, budget optimization, and performance assessment. - Apply advanced modeling techniques to quantify incremental impact across channels (sales force, digital, TV, congresses, patient programs). - Translate modeling outputs into clear business implications for brand strategy, media mix, and investment allocation. - Support annual brand planning, in‑year optimization, and scenario analysis using MMx outputs. From "Build & Operate Advanced MMx Capabilities": - Design and run advanced MMx models using Bayesian, Huber, etc. approaches, including: - Adstock and carryover modeling - Saturation and diminishing returns curves - Distributed lag and non-linear response functions - Cross‑channel interaction effects - Develop scalable and repeatable MMx pipelines aligned with enterprise standards and automated platforms. - Partner with centralized teams to integrate MMx into self‑service and agent‑enabled analytics solutions. - Ensure models are explainable, auditable, and fit for use in a regulated pharmaceutical environment. From "Agentic & Always‑On Insights Delivery": - Contribute to the design and execution of agentic and always‑on insights capabilities that provide continuous decision support to Brand, Omnichannel, and Field teams. - Operationalize MMx, KDA, and causal models into always‑on analytics pipelines with defined refresh cadence, data quality checks, and monitoring. - Structure model outputs (drivers, sensitivities, scenarios, constraints) to be consumable by AI‑enabled decision tools and agent‑based workflows. - Support “what changed / why” and forward‑looking recommendation logic using causal and driver‑based methods. - Ensure transparency, lineage, and explainability from data → model → insight → recommendation, with appropriate human‑in‑the‑loop review. - Define and maintain the Key Business Questions (KBQ) framework and business context layer, bridge business intent with the underlying technical semantic layer. - Partner with centralized platform, engineering, and Hyderabad teams to define inputs, features, and guardrails for agent‑enabled insights. - Design insights to be role‑based, actionable, and embedded into brand planning, optimization, and omnichannel execution workflows. - Drive adoption of agentic AI tools through role-based training, and structured feedback; serve as the primary business point of contact and channel user feedback into roadmap iteration. From "Governance, Compliance & Quality": - Adhere to data privacy, AI governance, and model validation standards (HIPAA, GDPR/CCPA, internal AI controls). - Maintain clear documentation of assumptions, methods, and limitations. - Participate in model reviews and audits with Legal, Privacy, and Governance stakeholders. From "Stakeholder Partnership": - Act as a trusted analytics partner to Brand, Marketing, Omnichannel, Field, and Finance teams. - Present insights in concise, business‑first narratives suitable for brand leadership and planning forums. - Support TA Analytics Leads by sharing best practices, templates, and modeling approaches. We need to keep sub-bullets? The rule: use
- for any list of duties/requirements/benefits. Could nest lists? Probably okay but simpler to flatten. However sub-items like under "Design and run advanced MMx models using Bayesian, Huber, etc. approaches, including:" then list of four items. We could keep as separate
- items, maybe include the leading phrase as part of li. Eg: "Design and run advanced MMx models using Bayesian, Huber, etc. approaches, including: Adstock and carryover modeling, Saturation and diminishing returns curves, Distributed lag and non-linear response functions, Cross‑channel interaction effects." Could combine. But better to keep each as separate li for clarity.
We'll produce a flat list; each bullet is a responsibility.
Qualifications: list items from qualifications section.
Extract qualifications bullet points:
- Advanced degree in Statistics, Economics, Data Science, Operations Research, or a related quantitative field.
- Minimum 5 years of experience in pharmaceutical commercial analytics, with direct ownership of MMx / MMM and causal analysis, or advanced ML models.
- Proven experience supporting brand strategy, budget planning, and performance measurement.
- Hands‑on expertise in advanced Marketing Mix Modeling, including:
- Bayesian and hierarchical modeling frameworks
- Adstock, lag, and saturation modeling
- Non‑linear optimization and scenario simulation
- Strong background in causal inference and incrementality measurement, such as:
- Geo‑experiments and matched‑market tests
- Synthetic control methods
- Uplift / incremental response modeling
- Proficiency in Python and/or R for statistical modeling, feature engineering, and automation.
- Solid SQL skills and experience working with large, multi‑source pharma datasets.
- Experience with cloud analytics platforms (e.g., Databricks, Snowflake) and production data pipelines.
- Familiarity with MLOps, version control, and model lifecycle management.
- Ability to translate analytical requirements into agent specifications, KBQ frameworks, and context layer definitions
- Experience leading or executing UAT for analytics or AI-enabled platforms, including test plan design, execution, and stakeholder coordination
- Ability to explain complex models and uncertainty to non‑technical stakeholders.
- Comfortable working with ambiguity, able to make sound analytical and business judgments when problem definitions, data, or requirements are not fully defined
- Product-minded analytical thinker, able to bridge the gap between analytical rigor and user experience design
- Domain & Platform Knowledge: Strong understanding of pharma commercial data (claims, APLD, specialty pharmacy, promotional, and digital engagement data).
- Experience integrating analytics with BI tools and decision workflows.
- Exposure to AI‑enabled or automated analytics platforms is a plus (agentic systems not required but beneficial).
- Demonstrated ability to drive end-user adoption of new analytics or AI tools through training, enablement, and change management
We'll flatten similarly.
Benefits: paragraph from Benefits section.
We need to include the benefits description: "Subject to the terms and conditions of the applicable plans then in effect, you may be eligible to participate in our comprehensive benefit plans – including wellbeing support, retirement and financial protection benefits, and insurance offerings (medical, dental, vision, life and disability). U.S.-based exempt employees are eligible for Flexible Time Off (FTO), which provides paid time off without a set accrual limit, subject to manager approval, along with 11 paid company holidays each year. Non-exempt employees, RayzeBio employees, and employees located in Puerto Rico receive 160 hours of paid vacation annually for new hires (subject to manager approval), 11 paid company holidays, and 3 optional holidays. Depending on eligibility, employees may also have access to additional time-off benefits, including paid sick leave, up to two paid volunteer days per year, summer hours flexibility, and leaves of absence for medical, personal, parental, caregiver, bereavement, or military needs. Eligible employees also enjoy an annual Global Shutdown between Christmas Day and New Year's Day. U.S.-based job seekers can explore full benefit offerings at https://careers.bms.com/benefits"
We'll keep as
. Pay: from Compensation Overview: "Princeton - NJ - US: $167,540 - $203,013 The starting pay range(s) listed above is for full-time employees (FTE). You may also be eligible for additional discretionary incentive cash and stock opportunities. We determine starting pay thoughtfully – carefully considering the nature of the role, required skills, work location, schedule and the knowledge and experience you bring. Final compensation is guided by pay equity principles and applicable employment laws. Compensation programs are reviewed on an ongoing basis and may be adjusted over time to reflect evolving market factors, and individual, team or Company performance." We'll put that in
under Pay heading. We should not include schedule section because none. Now produce HTML:
About the role
...summary...
Responsibilities
- ... each responsibility ... ...
Qualifications
- ... each qualification ... ...
Benefits
... benefits paragraph ...
Pay
... pay paragraph ...
Make sure no extra whitespace? It's fine. Now craft the list items. Responsibilities list items (I'll write them concisely but preserve meaning): 1. Own end‑to‑end development and execution of advanced Marketing Mix (MMx) and Key Driver Analyses (KDA) to support brand planning, budget optimization, and performance assessment. 2. Apply advanced modeling techniques to quantify incremental impact across channels (sales force, digital, TV, congresses, patient programs). 3. Translate modeling outputs into clear business implications for brand strategy, media mix, and investment allocation. 4. Support annual brand planning, in‑year optimization, and scenario analysis using MMx outputs. 5. Design and run advanced MMx models using Bayesian, Huber, etc. approaches, including: Adstock and carryover modeling, Saturation and diminishing returns curves, Distributed lag and non-linear response functions, Cross‑channel interaction effects. 6. Develop scalable and repeatable MMx pipelines aligned with enterprise standards and automated platforms. 7. Partner with centralized teams to integrate MMx into self‑service and agent‑enabled analytics solutions. 8. Ensure models are explainable, auditable, and fit for use in a regulated pharmaceutical environment. 9. Contribute