2 days ago
Warsaw, Poland or Kraków, PolandStaff+
Responsibilities
- Set ML architecture, modeling standards, validation practices, monitoring standards, and responsible-AI/model-risk practices for credit-scoring and merchant-underwriting models.
- Own the end-to-end model lifecycle from feature selection and training methodology through validation, deployment, and post-production drift and stability monitoring.
- Define point-in-time-correct production features with Data Engineering and shape the feature-store roadmap around modeling needs.
- Establish model documentation, model cards, adverse-action reasoning, and regulatory defensibility standards.
- Design and build agentic harnesses and loop/graph workflows for model iteration, evaluation, retraining, promotion, monitoring, and observability.
- Design reliable, observable, and cost-controlled agent loops and graphs covering context, tools, state, evaluations, and guardrails.
- Set technical direction, review designs and code, and mentor Senior and Mid ML Engineers.
- Partner with Credit Risk and engineering teams integrating models into decisioning services.
Requirements
- 7+ years of experience in machine learning engineering for production systems.
- 3+ years of experience in credit scoring, risk modeling, or decisioning ML using Spark, Databricks, or an equivalent lakehouse.
- 7+ years of expert Python and machine-learning framework experience with scikit-learn, XGBoost, and PyTorch.
- Strong experience designing credit-risk and underwriting models, including scorecards, gradient boosting, and survival or behavioral models.
- Experience with back-testing, champion/challenger testing, population stability, calibration, point-in-time-correct feature engineering, feature-store consumption, and MLflow-based MLOps.
- Fluency in explainability and responsible-AI techniques for regulated decisioning, including SHAP, model cards, and adverse-action reasoning.
- Working knowledge of regulatory model-risk frameworks and experience with agentic coding tools, agentic harnesses, loop and graph engineering, evaluations, guardrails, and iteration.
- Deep knowledge of LLM behavior and trade-offs.
- Staff-level technical leadership through influence, code and design review, and mentoring.
- English proficiency at B2 or higher.
- Fintech, lending, payments, fraud-detection modeling, real-time model serving, regulator-facing documentation, causal inference, and uplift-modeling experience are advantageous.
Benefits
- Hybrid work arrangement, ideally one day per week in the Warsaw or Kraków office.
- B2B contract.
- Architectural authority from day one on a greenfield, high-stakes fintech platform.
- Modern stack, lean AI-native team, autonomy, and opportunities to experiment with new technologies and bring them into production.
Tech Stack
Categories
About InPost
InPost is a public logistics company listed on Euronext Amsterdam that builds and operates an out-of-home parcel delivery network for e-commerce. Its network includes more than 50,000 out-of-home pick-up points, including over 27,000 parcel lockers, plus to-door courier and fulfillment services for online merchants across Poland, the UK, Italy, Iberia, France and Benelux. Headquartered in Luxembourg, it acquired French operator Mondial Relay in 2021.
