Customer Data Science Engineer, B2B Retention

What's The Role Liberty Business is building an advanced, data-driven customer retention capability across its Caribbean B2B markets. We are looking for a Customer Retention Science Engineer to lead the development of predictive analytics and data science capabilities that help the business identify customers at risk of churn, understand the drivers behind that risk, and translate analytical insights into actionable retention strategies. This role will own the B2B churn analytics capability end to end — from identifying and preparing the relevant data, defining churn targets and analytical populations, engineering features and developing predictive models, through to model validation, deployment, monitoring and business adoption. The role is intentionally broader than a traditional Data Scientist position. The successful candidate will have strong data engineering capabilities and will be expected to build and maintain the analytical datasets, data transformations and automated processes required to support predictive models and retention analytics. The primary focus, however, is Data Science and business impact, rather than infrastructure. The role should connect data, analytics and commercial decision-making, ensuring that models and insights ultimately contribute to reducing customer churn, protecting MRR and increasing customer lifetime value. You will work closely with the Retention & Loyalty team, BI, Commercial Operations, and local market teams to transform customer data into actionable intelligence.

What You'll Do
• Predictive Analytics & Churn Modeling:
• Own the development of the B2B customer churn prediction capability from data preparation through production and continuous improvement.
• Define appropriate churn targets, observation windows and performance windows in partnership with the Retention & Loyalty team.
• Build predictive models to identify customers with an increased probability of voluntary churn within a defined future period.
• Develop interpretable models that allow the business to understand why a customer is considered at risk, not only the probability of churn.
• Establish appropriate model validation frameworks, including train/test and out-of-time validation, ROC/AUC, Gini, KS and population stability monitoring.
• Identify and prevent temporal leakage and other sources of model bias or invalid predictive performance.
• Continuously improve the predictive capability as new customer, product, zbehavioral and service data becomes available.
• Feature Engineering & Customer Analytics
• Build and maintain customer-level analytical datasets and leak-safe learning sets that combine relevant B2B data sources.
• Engineer predictive variables from billing, contracts, products/RGUs, MRR, usage, service and technical incidents, customer care interactions, debt, customer experience and other relevant sources.
• Analyze customer behavior and identify patterns associated with churn, downgrades, contraction, renewal risk and changes in customer value.
• Develop customer segments and risk profiles that can be operationalized by Retention teams.
• Identify opportunities for proactive retention, cross-sell, upsell and customer value protection.
• Retention Strategy & Business Translation
• Translate analytical findings into clear business recommendations for Retention, Commercial and market leadership.
• Explain to non-technical stakeholders why a customer or segment is considered at risk and which factors are contributing to that risk.
• Partner with Retention Analysts and specialists to translate churn scores and insights into actionable intervention strategies.
• Support the development of next-best-action approaches based on customer risk, value, churn drivers and available retention treatments.
• Analyze the effectiveness of retention interventions and use results to improve future targeting and model performance.
• Measure the commercial impact of analytical initiatives, including churn reduction, MRR protected, customer value and retention treatment effectiveness.
• Data Engineering & Analytical Data Foundation
• Build and maintain the data pipelines, transformations and analytical datasets required to support churn modeling and retention analytics.
• Integrate data from multiple sources, including CRM, billing, contracts, customer care, network/service performance and customer experience platforms.
• Develop automated data preparation processes to reduce manual analytical work and improve reliability.
• Apply data validation and quality controls to ensure that analytical outputs are accurate, consistent and traceable.
• Troubleshoot data availability, integrity and transformation issues affecting analytical outputs.
• Collaborate with enterprise Data Engineering and BI teams when broader infrastructure or productionization support is required
• Model Governance, Documentation & Reproducibility
• Maintain clear documentation of data definitions, features, modeling methodology, assumptions, thresholds and limitations.
• Document model limitations and communicate them proactively to business stakeholders.
• Establish monitoring and governance practices that keep predictive models trustworthy over time.
• Maintain clear lineage and traceability from source data to analytical output and business action.
• Data Science Capability Building
• Raise the data science and analytical maturity of the Retention & Loyalty organization.
• Coach and support Retention Analysts in interpreting predictive outputs and applying analytical insights.
• Promote data-driven decision-making across the B2B retention organization.
• Stay current with developments in predictive analytics, machine learning, customer analytics and telecom retention practices. Tools & Technologies Core
• SQL: Advanced SQL for data extraction, transformation, analysis and feature engineering.
• Python: pandas, NumPy, scikit-learn, statsmodels and related analytical libraries.
• Data & Analytics: Experience working with large customer-level datasets and analytical data structures.
• AWS: S3, Athena and/or equivalent cloud data services.
• Data formats: Parquet and other analytical data formats.
• BI: Power BI or equivalent business intelligence platforms.
• Version control: Git or equivalent. Data Engineering
• ETL/ELT processes and data pipeline development.
• Relational databases and analytical data warehouses.
• Data modeling and data quality practices.
• API/data-source integration.
• Experience with cloud data platforms such as AWS Redshift, BigQuery or Azure Synapse is desirable.
• Advanced / Desired
• Machine learning and model deployment.
• Model monitoring and ML lifecycle management.
• Spark or other distributed data-processing technologies.
• Experience working with automated analytical pipelines. Mandatory:
• Bachelor’s degree in Statistics, Economics, Actuarial Science, Mathematics, Engineering, Computer Science, Data Science or another quantitative discipline.
• 3+ years of experience in data science, predictive analytics, customer analytics or a closely related field.
• Demonstrated experience building predictive models from raw data through validation and production or operational use.
• Strong understanding of statistical modeling and machine learning techniques.
• Advanced SQL skills and the ability to independently build analytical datasets.
• Strong Python skills for data preparation, feature engineering and predictive modeling.
• Strong understanding of temporal leakage and the design of time-based analytical datasets.
• Experience working with large and complex datasets from multiple source systems.
• Ability to independently investigate data issues and determine whether a problem is related to data quality, business definitions, transformation logic or model behavior.
• Ability to communicate complex analytical concepts clearly to non-technical stakeholders.
• Ability to translate analytical findings into business recommendations and measurable outcomes.
• Professional working proficiency in English. Desired:
• Experience building churn, propensity, risk, customer lifetime value or other customer-level predictive models.
• Experience in telecommunications, subscription businesses or B2B environments.
• Understanding of telecom data structures including billing, MRR, RGUs, contracts, usage, network/service performance and CRM.
• Experience deploying analytical outputs into BI or operational tools.
• Experience with AWS and cloud-based data environments.
• Experience with customer segmentation and next-best-action methodologies.
• Experience measuring the effectiveness of retention interventions or commercial treatments. Skills & Abilities
• Business-oriented data scientist: understands how analytical work translates into customer, revenue and retention outcomes.
• End-to-end ownership: able to move from a business problem to data, analysis, model, insight and operational action.
• Analytical curiosity: investigates why customer behavior changes rather than simply reporting what happened.
• Intellectual honesty: clearly communicates model limitations, assumptions and uncertainty.
• Strong problem-solving: able to troubleshoot both analytical and data-related problems independently.
• Commercial mindset: understands MRR, customer value, churn, retention economics and the importance of prioritizing high-value customers.
• Communication: able to explain technical decisions to executives and commercial stakeholders without oversimplifying them.
• Collaboration: comfortable working across Retention, BI, Data Engineering, Commercial Operations and local markets.
• Data discipline: strong focus on data quality, accuracy, reproducibility and traceability.
• Continuous learning: stays current with data science, machine learning, analytics and customer retention practices. SUCCESS IN THIS ROLE Success in this role will not be measured only by the number of models developed or datasets created. The successful candidate will be expected to demonstrate measurable progress in the organization’s ability to:
• Identify customers at risk of churn before the customer disconnects.
• Understand and explain the main drivers of customer churn.
• Prioritize retention efforts based on customer risk and commercial value.
• Provide Retention teams with actionable customer-level insights.
• Improve the effectiveness of retention interventions.
• Protect MRR and reduce avoidable customer churn.
• Build a scalable and sustainable predictive analytics capability across Caribbean B2B markets. The ultimate objective is to create a closed-loop retention intelligence capability in which customer data is transformed into predictive insight, predictive insight drives targeted intervention, and the results of those interventions continuously improve the organization’s understanding of customer behavior.