Senior Manager, Data Scientist

prompt • New York, NY

112d

About The Position

We are hiring a Senior Manager, Data Science to design and deliver the analytical work that powers measurement, audience strategy, and AI-augmented products across our pharma and consumer client portfolio. This is a hands-on role: you will write Python every day, pull data from APIs, build models, and turn the output into something a strategist or client can act on tomorrow morning.
You will work across structured client and third-party data (claims, syndicated panels, ad platforms, CRM) and unstructured signal (social, search, earned media, survey verbatims, internal documents). You will partner with strategy, creative, media, and engineering to move work from a business question to a defensible analytical answer, and you will help set the technical bar for the analytics team as it grows.
The right person is equally comfortable in a Jupyter notebook and in a client meeting. They have strong opinions about how to measure things, hold themselves to a high standard for analytical rigor, and can explain a model to a non-technical audience without dumbing it down or dressing it up.

Requirements

• 3–6 years of experience in data science, marketing analytics, or quantitative consulting, with at least 3 years writing Python in a professional setting.
• Python proficiency (non-negotiable): fluent in pandas, numpy, scikit-learn, and statsmodels; comfortable in Jupyter and at the command line; comfortable with virtual environments, git, and basic software hygiene.
• APIs and data engineering basics: you have built scripts that authenticate against REST APIs, paginate through results, handle rate limits and errors, parse JSON, and persist results to a structured store. Familiarity with webhook patterns and async requests is a plus.
• SQL: strong working knowledge — joins, window functions, CTEs, query optimization — across at least one major dialect (Postgres, BigQuery, Snowflake, Redshift).
• Statistics and modeling: working command of regression, classification, clustering, and time-series methods; understanding of train/test discipline, regularization, and validation; familiarity with causal inference, experimentation, or hierarchical models is a strong plus.
• Measurement experience: direct work on at least one of marketing mix modeling, multi-touch attribution, media measurement, brand/reputation measurement, or audience targeting or segmentation in a client-services or in-house marketing context.
• LLM tooling: practical experience with text data (cleaning, embeddings, classification, topic modeling) and with the major LLM APIs (Anthropic, OpenAI). Hands-on experience with RAG, agentic workflows, or LLM-powered tooling is preferred.
• Visualization and reporting: Plotly or matplotlib in Python. Able to design a chart that tells the truth and a deck that holds up under client scrutiny.
• Marketing data fluency: familiarity with social listening platforms (Brandwatch, Meltwater, Sprinklr, Talkwalker) or native social media data and/or social media APIs, digital analytics (GA4, Adobe), search/SERP data, and ad-platform reporting (Meta, Google, LinkedIn). Pharma-specific experience with claims, NPI data, or syndicated panels is a strong plus but not required.
• Communication: you can lead a working session with a client, present model results without hiding behind jargon, and write an email that does not need a follow-up.

Nice To Haves

• Experience in pharma, healthcare, or other regulated industries.
• Cloud platform experience (AWS, GCP, or Azure) at the level of running notebooks, storing data, and deploying small services.
• Experience shipping lightweight applications — Streamlit, Gradio, FastAPI, Flask — to put a model or tool in front of internal users or clients.
• Familiarity with Bayesian methods, hierarchical models, or modern MMM frameworks (PyMC, Robyn, Meridian, LightweightMMM).
• Published writing, open-source contributions, or a portfolio of analytical work.
• Curious and rigorous. You ask why the data looks the way it does before you start modeling. You know the difference between a result and an artifact, and you would rather say "I don’t know yet" than overclaim.
• Owner, not a passenger. You drive your own work. You follow up without being chased. You flag risks early. You finish things.
• Bilingual in technical and non-technical. You can defend a methodological choice to another data scientist and explain the same choice to a brand lead in a way that lands.
• Teacher and teammate. You enjoy mentoring junior people, you give feedback honestly and kindly, and you make the people around you better.
• Comfortable with ambiguity. Agency work is fast, the data is rarely clean, and the answer is rarely waiting for you in a textbook. You are okay with that.

Responsibilities

• Build and own analytical models for client work — marketing mix and multi-touch attribution, audience segmentation, propensity and uplift models, time-series forecasting, and measurement frameworks for earned, owned, and paid media.
• Write production-quality Python for data pipelines, modeling, and reporting. Pull from REST APIs (social listening, search, ad platforms, syndicated data vendors, internal tools), handle auth, pagination, rate limits, and schema drift, and ship code others on the team can read and extend.
• Work with messy, real-world data — claims and NPI-level data, social and search corpora, survey results, CRM exports, ad-platform reports — and turn it into clean, joined, model-ready datasets.
• Apply NLP and LLM tooling to text data: classification, topic modeling, embeddings, and retrieval-augmented workflows over social posts, earned media, survey verbatims, and internal knowledge bases. Prototype LLM-powered features for client deliverables and internal products.
• Translate findings into client-ready outputs — decks, dashboards, written narratives, recommendations — that are honest about uncertainty and clear about what the client should do next.
• Partner across the agency with Strategy, Planning, Creative, Engagement, and Digital to scope analytical work that actually answers the business question, and to bring measurement thinking into the front end of campaigns rather than the end.
• Mentor analysts and junior data scientists on Python, statistics, modeling craft, and how to communicate quantitative work to a non-technical audience. Raise the floor on code review, documentation, and reproducibility.
• Contribute to internal tools and IP — measurement frameworks, reusable code libraries, dashboards, and AI-augmented internal products — that compound the team’s capability over time.
• Run multiple workstreams in parallel, manage timelines and deliverables, and keep clients and internal stakeholders informed without being asked.

Benefits

• Provide a safe and respectful workplace, that is diverse and inclusive in nature
• Always be available to listen to your needs, concerns, and feedback
• Help you articulate and achieve your goals
• Purposefully drive business that meets your passions
• Share best practices, industry news and innovations
• Give you the time you need to have wellness and mindfulness

What This Job Offers

• Job Type: Full-time
• Career Level: Senior
• Education Level: No Education Listed