Annapurna are proud to be representing Banxware – Banxware is Europe’s leading embedded finance provider, empowering platforms to offer tailored lending solutions to small and medium-sized businesses (SMEs). We partner with a wide range of platforms where SMEs generate revenue, including payment and POS providers, eCommerce marketplaces, and shop software vendors. Additionally, we collaborate with platforms aggregating multiple SMEs, such as neo banks, logistics providers, cash management platforms, and more. Our innovative solutions are trusted by top-tier companies like Qonto, Deutsche Bank (FYRST), Agicap, and JustEat Takeaway (Lieferando). From day one, we’ve operated on an international scale, and we’re proud to embrace a global mindset in everything we do.
Berlin – Hybrid (2 days in office)
About the role
We’re hiring a Senior Data Scientist to lead analytics-driven decisioning across core business functions. This role has three primary focuses: business analytics and strategic insight, risk and portfolio analytics (last two highly desirable but not necessary), and building production-grade ML/AI solutions – including MLOps and LLM-based features. You’ll partner closely with our Risk-focused data science team and engineering partners to turn complex data into reliable, scalable AI products that drive measurable business outcomes.
What you’ll do (key responsibilities)
* Lead end-to-end analytics and modeling projects that deliver actionable business insights for stakeholders (commercial, pricing, operations, credit, etc.).
* Design, develop, validate, and deploy advanced statistical models and machine learning solutions for business metrics (propensity, churn, forecasting, CLTV, segmentation, etc.).
* Support building and productionizing models and AI systems using modern MLOps practices: CI/CD for ML, model versioning, monitoring, automated retraining, and scalable serving.
* Architect and implement LLM and other generative AI solutions safely and effectively in production (prompt engineering, fine-tuning/adapter strategies, retrieval-augmented generation, hallucination mitigation, guardrails).
* Collaborate with the Risk & Portfolio analytics team to extend models for risk assessment, portfolio health, stress testing, and provisioning. Translate risk requirements into robust model assumptions and validation steps.
* Establish and maintain model governance, documentation, testing, and observability aligned with enterprise standards and regulatory expectations.
* Upskill Banxware’s data capabilities including other data scientists and engineers
* Communicate technical findings and trade-offs clearly to senior stakeholders and non-technical partners; recommend measurable business actions.
* Lead pilots and proofs-of-concept; convert successful pilots into production-grade systems in partnership with engineering teams.
Required qualifications
* 5+ years experience in applied data science, machine learning, or analytics with demonstrable end-to-end project ownership.
* Strong experience in business analytics (churn, LTV, sales/marketing analytics, forecasting, experimentation).
* Proven track record deploying machine learning models into production using MLOps practices (CI/CD, model serving, monitoring).
* Knowledge implementing LLM-based solutions in production (prompting, fine-tuning, RAG architectures, safety/guardrails).
* Hands-on with common ML/AI stack: Python, scikit-learn, TensorFlow/PyTorch, experience with containerization (Docker) and orchestration.
* Solid SQL skills and experience working with production data platforms (cloud data warehouses, data lakes).
* Familiarity with model governance, testing strategies, and performance monitoring.
* Ability to translate analytic results into business recommendations and influence senior stakeholders.
Desirable but not necessary
* Prior experience in Fintech
* Knowledge of regulatory model requirements (model validation, explainability, fairness, documentation) is a plus.
* Experience with feature stores, model registries (MLflow, Sagemaker), and inference serving platforms.
* Familiarity with cloud platforms (in specific AWS) and data engineering patterns (ETL/ELT, streaming).
* Familiarity with experimentation platforms and causal inference methods.
Call Settings Override
