Data Scientist (Manager)

Company description Publicis Groupe is the largest Communications Group worldwide and the leader in Digital and Interactive Communications. Publicis has activities spanning 108 countries on five continents and employs approximately 72,000 professionals worldwide. Publicis Groupe offers local and international clients a complete range of communication services through the nearly 1,400 agencies across our four global networks, including: Publicis Re:Sources Re:Sources is the shared services provider of Publicis Groupe, delivering a suite of multi-tenant managed and professional services to Publicis Groupe agencies worldwide, in support of key Groupe business operations. Those operations include: Information Technology & Technology Solutions, Finance, Legal, Procurement, Real Estate, Insurance and other services to our business units Overview We are seeking a highly skilled and experienced Data Scientist (Manager) to lead advanced analytics and machine learning initiatives within our organization. This role will be responsible for driving data-driven decision-making, building scalable AI/ML solutions, and managing a team of data scientists to deliver impactful business outcomes. As a Manager, you will play a critical role in bridging business objectives with technical solutions, partnering with cross-functional teams to design, develop, and deploy innovative data science models. The ideal candidate will bring strong expertise in statistical modeling, machine learning, and data engineering, along with proven leadership experience in managing high-performing teams. Responsibilities 1) Strategy, Roadmap & Stakeholder Management Own the multi‑quarter data science roadmap aligned to product/BU OKRs; define measurable success criteria and drive adoption of AI capabilities. Translate ambiguous business problems into ML problem statements, prioritize opportunities using impact/effort and data readiness. Lead make/buy and tool selection decisions; run PoCs and establish guardrails for scale. 2) Team Leadership & Talent Lead and grow a high‑performing team (typically 6–12 DS/ML/Applied DS), including hiring, mentoring, career development and performance management. Foster a culture of scientific rigor, engineering excellence and responsible innovation; establish a Community of Practice. 3) Technical Direction & Architecture Define reference patterns for the ML lifecycle: data prep, feature stores, experimentation, model registry, deployment and monitoring across Azure/AWS/GCP. Set standards for reproducible research (tracking, seeds, data versioning), governed features and model documentation (model cards, datasheets). Guide choices among classical ML, deep learning, NLP/LLMs and rules/heuristics based on constraints and business value. 4) Delivery & MLOps Oversee delivery of end‑to‑end ML solutions: robust features/pipelines, offline/online inference paths and reliable rollouts (blue/green, shadow, canary). Enforce CI/CD for ML, automated evaluations, bias/fairness checks and model registry workflows; ensure DR/BCP for critical models. Partner with Data/Platform/DevOps teams to productionize solutions using MLflow/W&B, Databricks/Vertex/SageMaker/Azure ML, Airflow/ADF/Composer, Terraform/Bicep. 5) Responsible AI, Governance & Risk Implement model governance: lineage, approvals, documentation and audit trails; manage model risk with periodic reviews and challenger models. Operationalize explainability, fairness, drift and performance monitoring; define alert thresholds and retraining policies. Ensure compliance with privacy, security and regulatory standards (e.g., GDPR/CCPA, sector guidelines) and enterprise data policies. 6) Analytics, Insights & Communication Champion evidence‑based decisioning: statistical analysis, experimentation design and causal methods where appropriate. Build executive‑ready narratives and dashboards that connect model outputs to business KPIs, ROI and risk. 7) Cost, Reliability & Operations Own availability, latency and accuracy SLAs/SLOs for production models; lead incident response and postmortems. Drive FinOps for ML workloads (training/inference cost, GPU utilization, auto‑scaling) and optimize time‑to‑value. Qualifications Must have skills 10+ years in data science/ML (or closely related), including 3+ years of people leadership (hiring, coaching, performance management). Bachelor’s/Master’s in Computer Science, Data Science, Mathematics, Statistics, Engineering, or related field. Strong programming in Python and practical expertise with pandas, NumPy, scikit‑learn; working knowledge of TensorFlow or PyTorch. Proven experience developing, tuning and validating supervised/unsupervised models (classification, regression, clustering, time series/forecasting), including robust feature engineering, EDA and statistical analysis. Solid understanding of ML algorithms, probability, linear algebra, optimization and statistical inference; capable of guiding method selection and experimental design. Experience with large-scale datasets and distributed computing (Spark, Dask) and/or cloud ML platforms: Azure ML, Databricks ML, GCP Vertex AI, AWS SageMaker Proficiency with SQL for complex data extraction, joins, window functions and transformations. Practical MLOps skills: experiment tracking (MLflow, Weights & Biases), model versioning/registry, reproducibility, automated testing, CI/CD for ML and deployment pipelines. Experience deploying models to production via REST APIs, containers (Docker), serverless (e.g., Azure Functions, AWS Lambda, Cloud Run), or batch/streaming inference. Understanding of model monitoring, drift detection, data/label quality and explainability (SHAP, LIME, counterfactuals); exposure to model governance and risk management processes. Strong communication and storytelling able to translate findings into decisions using visualizations (Matplotlib/Seaborn, Plotly) and BI tools (Power BI, Tableau, or similar). Nice to have skills- NLP & LLMs: transformers, embeddings, vector databases (e.g., Pinecone, Weaviate, Milvus, Redis), retrieval‑augmented generation (RAG), prompt engineering, evaluation frameworks. Anomaly detection, recommender systems, uplift modeling, causal inference, experimentation (A/B testing, CUPED). Experience with Databricks (Delta/Unity Catalog), Snowflake (Snowpark ML), BigQuery ML, Azure ML pipelines. Containers & orchestration: Kubernetes (AKS/EKS/GKE), Kubeflow Pipelines, Ray, microservice-based ML deployment. Feature stores (Feast, Databricks Feature Store, SageMaker Feature Store) and streaming inference (Kafka, Kinesis, Pub/Sub). Privacy‑preserving ML (data minimization, PII handling, differential privacy, federated learning), domain knowledge in BFSI/Healthcare/Retail. Additional information Work Schedule Core work hours are Monday through Friday, 9:00 AM-5:00 PM. Must also be flexible and be available to work non-standard business hours upon request or as needed Must be available via cell phone for VIP support Travel Local travel between sites may be required Occasional travel to sites outside the local area may be required Salary Range: Transparency matters to us. The salary range for this position is $125000-$140000 per year. Actual compensation within this range will be based on a variety of factors, including relevant experience, knowledge, skills, and applicable certifications. This range reflects what we reasonably expect to offer based on current market data. Job description only reflects management’s assignment of essential functions. Management reserves the right to assign or reassign duties and responsibilities to this job at any time. This job description in no way states or implies that these are the only duties to be performed by the employee(s) currently in this position. Employee(s) will be required to follow any other job related instructions and to perform any other job-related duties requested by any person authorized to give instructions or assignments. A review of this position has excluded the marginal functions of the position that are incidental to the performance of fundamental job duties. All duties and responsibilities are essential job functions and requirements and are subject to possible modification to reasonably accommodate individuals with disabilities. To perform this job successfully, the incumbent(s) will possess the skills, aptitudes, and abilities to perform each duty proficiently. Some requirements may exclude individuals who pose a direct threat or significant risk to the health or safety of themselves or others. The requirements listed in this document are the minimum levels of knowledge, skills, or abilities. This document does not create an employment contract, implied or otherwise, other than an "at-will" relationship. Re:Sources USA is an Equal Opportunity / Affirmative Action employer. All qualified applicants to Re:Sources USA will receive consideration for employment without regard to race, color, religion, sex, national origin, disability, or protected veteran status.

Must have skills 10+ years in data science/ML (or closely related), including 3+ years of people leadership (hiring, coaching, performance management). Bachelor’s/Master’s in Computer Science, Data Science, Mathematics, Statistics, Engineering, or related field. Strong programming in Python and practical expertise with pandas, NumPy, scikit‑learn; working knowledge of TensorFlow or PyTorch. Proven experience developing, tuning and validating supervised/unsupervised models (classification, regression, clustering, time series/forecasting), including robust feature engineering, EDA and statistical analysis. Solid understanding of ML algorithms, probability, linear algebra, optimization and statistical inference; capable of guiding method selection and experimental design. Experience with large-scale datasets and distributed computing (Spark, Dask) and/or cloud ML platforms: Azure ML, Databricks ML, GCP Vertex AI, AWS SageMaker Proficiency with SQL for complex data extraction, joins, window functions and transformations. Practical MLOps skills: experiment tracking (MLflow, Weights & Biases), model versioning/registry, reproducibility, automated testing, CI/CD for ML and deployment pipelines. Experience deploying models to production via REST APIs, containers (Docker), serverless (e.g., Azure Functions, AWS Lambda, Cloud Run), or batch/streaming inference. Understanding of model monitoring, drift detection, data/label quality and explainability (SHAP, LIME, counterfactuals); exposure to model governance and risk management processes. Strong communication and storytelling able to translate findings into decisions using visualizations (Matplotlib/Seaborn, Plotly) and BI tools (Power BI, Tableau, or similar). Nice to have skills- NLP & LLMs: transformers, embeddings, vector databases (e.g., Pinecone, Weaviate, Milvus, Redis), retrieval‑augmented generation (RAG), prompt engineering, evaluation frameworks. Anomaly detection, recommender systems, uplift modeling, causal inference, experimentation (A/B testing, CUPED). Experience with Databricks (Delta/Unity Catalog), Snowflake (Snowpark ML), BigQuery ML, Azure ML pipelines. Containers & orchestration: Kubernetes (AKS/EKS/GKE), Kubeflow Pipelines, Ray, microservice-based ML deployment. Feature stores (Feast, Databricks Feature Store, SageMaker Feature Store) and streaming inference (Kafka, Kinesis, Pub/Sub). Privacy‑preserving ML (data minimization, PII handling, differential privacy, federated learning), domain knowledge in BFSI/Healthcare/Retail.

1. Strategy, Roadmap & Stakeholder Management Own the multi‑quarter data science roadmap aligned to product/BU OKRs; define measurable success criteria and drive adoption of AI capabilities. Translate ambiguous business problems into ML problem statements, prioritize opportunities using impact/effort and data readiness. Lead make/buy and tool selection decisions; run PoCs and establish guardrails for scale. 2) Team Leadership & Talent Lead and grow a high‑performing team (typically 6–12 DS/ML/Applied DS), including hiring, mentoring, career development and performance management. Foster a culture of scientific rigor, engineering excellence and responsible innovation; establish a Community of Practice. 3) Technical Direction & Architecture Define reference patterns for the ML lifecycle: data prep, feature stores, experimentation, model registry, deployment and monitoring across Azure/AWS/GCP. Set standards for reproducible research (tracking, seeds, data versioning), governed features and model documentation (model cards, datasheets). Guide choices among classical ML, deep learning, NLP/LLMs and rules/heuristics based on constraints and business value. 4) Delivery & MLOps Oversee delivery of end‑to‑end ML solutions: robust features/pipelines, offline/online inference paths and reliable rollouts (blue/green, shadow, canary). Enforce CI/CD for ML, automated evaluations, bias/fairness checks and model registry workflows; ensure DR/BCP for critical models. Partner with Data/Platform/DevOps teams to productionize solutions using MLflow/W&B, Databricks/Vertex/SageMaker/Azure ML, Airflow/ADF/Composer, Terraform/Bicep. 5) Responsible AI, Governance & Risk Implement model governance: lineage, approvals, documentation and audit trails; manage model risk with periodic reviews and challenger models. Operationalize explainability, fairness, drift and performance monitoring; define alert thresholds and retraining policies. Ensure compliance with privacy, security and regulatory standards (e.g., GDPR/CCPA, sector guidelines) and enterprise data policies. 6) Analytics, Insights & Communication Champion evidence‑based decisioning: statistical analysis, experimentation design and causal methods where appropriate. Build executive‑ready narratives and dashboards that connect model outputs to business KPIs, ROI and risk. 7) Cost, Reliability & Operations Own availability, latency and accuracy SLAs/SLOs for production models; lead incident response and postmortems. Drive FinOps for ML workloads (training/inference cost, GPU utilization, auto‑scaling) and optimize time‑to‑value.