About the Job
We are seeking a Data Scientist who will lead analytical design, model development, and experimental on across high-impact customer analytics initiatives such as Customer Lifetime Value (CLV), Predictive NPS/CSAT, Churn/Retention, and Customer Segmentation. The ideal candidate has strong hands-on expertise in predictive modelling, large-scale distributed computing (PySpark), and real-world deployment experience. A proven ability to work with large, complex datasets to translate business needs into scalable analytical solutions.
Your Main Duties Will Include:
Customer Analytics & Predictive Modelling
- Build and maintain CLV models (historical and/or predictive) incorporating revenue, costs, engagement signals, and churn/retention.
- Develop predictive NPS, satisfaction, and churn models to identify high-risk customers and key drivers of experience.
- Design customer segmentation (value-based, behavioral, RFM, clustering, predictive) to support targeting, campaigns, and product design.
- Ensure all models are robust, monitored, and explainable, with clear links to business objectives and measurable impact.
Data Wrangling & Feature Engineering
- Work with large, complex datasets from multiple sources such as CRM, transactional and interaction data, digital journeys, contact center data, and surveys/VoC.
- Use PySpark, Python, and SQL to clean, transform, and join data; build scalable feature pipelines.
- Partner with engineering to productionize models via batch jobs, APIs, and dashboards; ensure reliability and performance.
Evaluation & Explainability
- Define and track relevant metrics (AUC, F1, uplift, calibration, segment performance, stability, etc.).
- Use explainability techniques (feature importance, SHAP, or similar) to communicate model behavior clearly.
- Contribute to documentation, model monitoring, and retraining plans to sustain performance over time.
Stakeholder Engagement & Storytelling
- Translate business questions into clear analytical problems, hypotheses, and success criteria.
- Present insights and recommendations to CX, Marketing, Product, Digital, and Operations stakeholders.
- Prepare concise decks, summaries, and dashboards to support decisions and drive adoption.
Requirements:
Experience
- 4+ years of hands-on experience in Data Science and Predictive Modelling within established or high-growth organizations.
- Bachelor’s or Master’s degree in Data Science, Statistics, Mathematics, Computer Science, or a related quantitative field.
- Proven experience in end-to-end delivery of data science solutions—from problem framing and data preparation to model deployment and monitoring.
- Demonstrated delivery of at least one (production or advanced PoC) of the following:
- CLV / Profitability
- Churn / Propensity
- Predictive NPS / CSAT
- Customer segmentation at scale
Technical Skills (Must-Have)
- Strong proficiency in Python (pandas, scikit-learn, MLlib, or similar libraries).
- Strong proficiency in PySpark and working with large datasets on distributed platforms.
- Advanced SQL skills for complex querying, data transformation, and performance optimization.
- Experience using Power BI, Tableau, or similar tools for analytics, dashboards, and data storytelling.
- Experience with MLflow (or similar) for experiment tracking, model versioning, and lifecycle management.
- Solid understanding and hands-on experience with core Data Science concepts, including:
- Supervised learning: classification, regression, uplift modelling
- Unsupervised learning: clustering, dimensionality reduction
- Feature engineering, model tuning, and validation
- Experience working in a cloud or big data environment (Azure, AWS, GCP, Databricks, or similar).
Preferred Qualifications (Bonus)
- Experience with Experience Management platforms such as Medallia, Qualtrics, or similar.
- Strong practical MLOps habits (monitoring, drift checks, reproducible pipelines, CI/CD collaboration).
- Exposure to NLP and text analytics (e.g., survey verbatims, call-center transcripts, complaint logs) applied to customer experience or insight use cases.