Senior Data Scientist

This is a fully remote position, open to applicants in Romania.

📋 Description

• Take ownership of the entire lifecycle of intricate modelling programs within Customer & Commercial Intelligence: CLTV, Churn Prediction, Propensity to Buy, and Next Most Likely Product (NMLP).

• Design multi-horizon churn models and create the churn intervention scoring layer that prioritizes at-risk merchants for the Account Management teams.

• Direct the NMLP engine by designing and implementing a multi-output recommendation system that identifies the next product from myPOS's complete catalog.

• Assume technical responsibility for core fraud model components, including transaction-level classifiers, merchant behavior anomaly detectors, and new-account fraud scorers optimized for high-throughput, low-latency inference.

• Architect and manage the Next Best Action (NBA) decision-making engine, a real-time system that selects the action with the highest expected value for each merchant during every interaction.

• Design and develop production-grade agentic AI systems that automate valuable analytical and operational workflows.

• Define and implement experimental designs for online assessments, including A/B tests, uplift experiments, and bandits, and analyze the results with statistical rigor.

• Establish and uphold technical standards across the team, including code quality, reproducibility, evaluation rigor, model documentation, and MLOps practices.

• Create high-quality model documentation and effectively communicate complex modelling projects to stakeholders in Sales, Marketing, Risk, Product, and Operations.

⛳️ Requirements

• MSc or PhD in Computer Science, Statistics, Applied Mathematics, Econometrics, or a related quantitative discipline (or equivalent professional experience).

• Over 7 years of hands-on experience in applied data science and machine learning within a commercial context, along with a robust portfolio of models in production that have delivered measurable business results.

• Proficient in Python for data science and ML engineering: pandas, scikit-learn, XGBoost/LightGBM, PyTorch, or TensorFlow; consistently producing clean, tested, modular code.

• Extensive knowledge across the entire ML methodological spectrum, including survival analysis, time-series and sequence modeling, uplift and causal inference, anomaly detection, and recommendation systems.

• Demonstrated end-to-end ownership of at least three of the following: CLTV models, churn models, propensity models, fraud/risk models, recommendation systems, or NBA systems in a commercial production environment.

• Strong capabilities in MLOps, including feature stores, model registries, model serving infrastructure, drift monitoring, and CI/CD for ML pipelines.

• Proficient in SQL and data platforms (GCP/BigQuery preferred); experience in streaming architectures for real-time feature generation.

• Practical experience in building LLM-powered applications, including RAG pipelines, tool-use agents, multi-agent orchestration, and agent evaluation frameworks.

• Strong background in causal inference techniques such as uplift modeling, difference-in-differences, or instrumental variables.

• Exceptional communication skills, capable of presenting complex technical work to senior business stakeholders and producing high-quality model documentation.

🏝️ Benefits

• Dynamic international team working in a high-tech environment.

• Competitive compensation package.

• Access to myPOS Academy for professional development and training.

• Unlimited access to courses via LinkedIn Learning.

• Bonus for referring a friend, as we understand that working with friends makes it enjoyable.

• Opportunities for team building, social activities, and networking on a multinational scale.