Data Scientist

Quick Facts

• Role: Data Scientist

• Focus: Fraud detection and fraud risk modeling in an insurance transformation program

• Collaboration: Multidisciplinary team; hybrid work (1 day in Warsaw office, 4 days remote)

Description

You will help build a modern claims and benefits ecosystem using data, workflow automation, and AI. The role centers on developing and maintaining machine-learning models for fraud detection and fraud risk identification across the full lifecycle—from data and feature preparation to validation and production deployment. You will translate business problems into analytical tasks and conduct experiments while communicating results to technical and business stakeholders.

Responsibilities

• Build, develop, and maintain models assessing fraud and misuse risk

• Own the full ML lifecycle: data analysis, feature preparation, validation, and production deployment

• Convert business problems into analytical tasks and choose modeling methods for tabular, text, and graph data

• Design and run research experiments; evaluate outcomes

• Find new data sources and test innovative analytical methods

• Prepare documentation and explain analysis results to technical and business stakeholders

• Collaborate closely with domain experts and the development team during implementation

Requirements

• Minimum 3 years experience as a Data Scientist; participated in at least one project with a model deployed for real business use

• Ability to evaluate production model effectiveness and interpret and justify model decisions

• Excellent SQL

• Practical experience building classification models for tabular data (from logistic regression to gradient boosting)

• Experience in feature engineering

• Ability to select validation methods, metrics, and model quality assessment approaches

• Knowledge of model interpretability methods such as SHAP (or others)

• Ability to independently run experiments using programming tools

• Understanding of how data quality and data provenance impact analytical model performance

• Strong collaboration skills with business experts and ability to present to non-technical audiences

• Awareness of current ML trends and judgment on when using LLMs is justified by business needs

Benefits

• Cafeteria-style benefits

• B2B contract

• Hybrid model: at least 1 day per week in the Warsaw client office

• Work on a project with real impact on the client’s business direction