About the role
Xtremepush is seeking an experienced Data Analyst to maintain predictive models powering customer decisioning across multiple client projects. You will lay the data foundation for simple models today and more sophisticated modelling as data volume grows, working with business users and technical stakeholders.
The role requires strong analytical and problem-solving skills, plus excellent written communication, in a hybrid Dublin setting.
Own production models end to end, including churn, customer lifetime value (CLV), propensity and similar customer models. Validate performance against holdout data and live outcomes, and monitor models for drift, bias and degradation. Define the data structure, entities and attribute design across each client’s full customer dataset, and document the data requirements that our Data Modeller implements and maintains. Identify data gaps and quality issues, and agree fixes with client and internal teams. Document models, features and assumptions to a standard suitable for a regulated environment. Explain model results clearly to non-technical stakeholders, and feed learnings back to Product to help shape the product roadmap. 5+ years’ experience building, deploying and maintaining customer models in live environments, ideally in financial services or another regulated sector. A degree in a quantitative field such as statistics, computer science, mathematics or economics, or equivalent experience. Strong Python, including common ML libraries, and advanced SQL. Strong data modelling skills, including dimensional modelling and handling data that changes over time. Experience with ClickHouse or another columnar database. Solid grounding in model and data validation, such as holdout design and calibration. Experience defining metrics and building BI dashboards and the datasets behind them, in a tool such as Metabase. Experience with cloud data platforms such as AWS or Snowflake. Strong problem-solving skills and excellent written communication. Experience with customer engagement, CRM or marketing decisioning platforms, or with marketing and campaign data, is desirable. Experience with MLOps practices, such as experiment tracking, model registries and CI/CD for models, is desirable. Experience with dbt or a similar SQL modelling framework, using Git for version control, is desirable. Experience with MySQL as a source system is desirable. Hybrid role Office in Dublin