Data Scientist

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

📋 Description

• Develop and construct predictive models from the ground up, beginning with:

• Analyze raw trading, transactional, and behavioral data sourced from our data warehouse.

• Identify target variables and translate business concepts (e.g., defining "churn" within a brokerage context) into quantifiable ML objectives.

• Create features based on client interactions, trading behaviors, market conditions, and engagement indicators.

• Choose, train, validate, and refine models — starting with simple approaches and increasing complexity where it proves beneficial.

• Establish monitoring systems for model performance, data drift, and degradation over time.

• Provide daily client-level scores that seamlessly integrate into CRM workflows and sales operations.

• Convert model outputs into practical insights for non-technical sales managers.

• Collaborate with sales leadership to design interventions based on model predictions.

• Present findings, assumptions, limitations, and recommendations to senior stakeholders.

⛳️ Requirements

• Over 4 years of hands-on experience in building and deploying predictive models to address real business challenges (classification, regression, scoring).

• High proficiency in Python (pandas, scikit-learn, XGBoost/LightGBM/CatBoost) and SQL.

• Proven ability to independently frame ambiguous business challenges as ML tasks — defining targets, engineering features, and selecting methodologies.

• Experience managing tabular data at scale: feature engineering, addressing class imbalance, temporal validation, and preventing data leakage.

• Capability to communicate model results to non-technical stakeholders in clear, actionable terms.

• Background in working with time-series or event-based behavioral data.

• Experience with churn prediction, propensity modeling, CLV, or customer scoring across any industry (considerable advantage).

• Familiarity with survival analysis (Cox proportional hazards, time-to-event modeling) (considerable advantage).

• Experience in model monitoring within production environments: data drift detection, retraining processes, champion-challenger frameworks (considerable advantage).

• Experience in financial services, brokerage, or fintech (considerable advantage).

• Background in probabilistic models for CLV (BG/NBD, Pareto/NBD, Gamma-Gamma) (considerable advantage).

• Familiarity with SHAP, LIME, or other model interpretability methods (considerable advantage).

• Experience with data warehousing tools (BigQuery, Databricks, or similar) (considerable advantage).

🏝️ Benefits

• Health insurance

• 401(k) matching

• Flexible work hours

• Paid time off

• Professional development opportunities