> Markdown version of [/jobs/ext/4622593-director-data-science-and-machine-learning-risk-and-fraud](https://www.wearedevelopers.com/jobs/ext/4622593-director-data-science-and-machine-learning-risk-and-fraud). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Director, Data Science and Machine Learning (Risk and Fraud) - **Company:** Trustly Inc. - **Location:** New York, NY, United States (Remote available) - **Experience:** Experienced - **Contract:** Temporary contract - **Skills:** Amazon Web Services, Big Data, BigQuery, Cloud Database, Python (Programming Language), Link Analysis, Machine Learning, Standard Sql, Transaction Data, Feature Store, Snowflake, Apache Spark, Google Vertex AI, Pandas, Scikit Learn, Real-time Inference, Information Technology, Xgboost, Software Library, Amazon Redshift, Databricks - **Published:** November 7, 2026 - **Apply:** https://www.juju.com/job/28_25197_025301d77ac2e5c81684c7ae906ceb49-2362439735 ## About the Role * 8+ years of experience building and deploying machine learning models into production, including at least 2 years leading a team. * A leader who remains deeply hands-on and is currently capable of independently building, deploying, and evaluating a production machine learning model. * Strong Python and SQL skills, with experience working with large-scale datasets using distributed processing engines or cloud data warehouses such as Athena, Trino, Spark, Redshift, Snowflake, or BigQuery. * Direct experience in fraud, credit risk, payments risk, or another high-stakes decisioning environment, including rare-event modeling, label delay, reject inference, and understanding the difference between strong model metrics and strong business decisions. * Experience with real-time inference and production model monitoring in a low-latency decisioning environment, including drift detection, retraining strategies, and model updates. * Experience owning or materially contributing to a loss budget or loss forecast, with the ability to translate model behavior and risk performance into financial impact for executives and auditors. * Strong experimentation experience for model and policy changes, including champion/challenger testing, holdouts, staged rollouts, and rigorous measurement of outcomes. * Demonstrated experience mentoring and developing Data Scientists and Machine Learning Engineers., * Bachelor's degree in a quantitative discipline such as Statistics, Mathematics, Computer Science, Economics, Engineering, or Operations Research. * Strong proficiency with pandas and machine learning libraries such as scikit-learn, XGBoost, LightGBM, or CatBoost. * ACH or EFT domain expertise, including familiarity with NACHA return codes and return economics. * Experience working with open banking or bank transaction data. * Experience with cloud machine learning platforms such as Databricks, AWS SageMaker, or GCP Vertex AI. * Experience with feature stores, streaming feature computation, rules engines, or real-time decisioning systems. * Experience applying sequence or graph-based modeling techniques to behavioral or transaction data, such as recurrent architectures or entity-link analysis for identifying fraud networks. * Exposure to model risk management and governance within financial services. * Experience evaluating, integrating, or managing third-party risk data and model providers. ## Description You will lead the team responsible for building and deploying the machine learning models that determine whether transactions are approved and guaranteed. This is a hands-on technical leadership role. You will build and ship models yourself while leading, mentoring, and developing a team of Data Scientists and Machine Learning Engineers. You will own not only the team's roadmap, but also the quality, performance, and business impact of what it delivers in production. Your work will sit at the intersection of machine learning, payments risk, product, and engineering, with direct impact on approval rates, fraud and loss performance, customer experience, and Trustly's broader payments network., * Build, validate, deploy, and continuously improve machine learning models across ACH return risk, fraud, and account takeover. * Develop real-time, online, and offline features using transaction, account, device, and behavioral data. * Design and analyze experiments to evaluate model and policy changes, including champion/challenger tests, holdouts, and staged rollouts. * Quantify the impact of model changes across loss rates, approval rates, ROI, customer experience, and merchant outcomes. * Translate complex model performance and findings into clear, actionable insights for non-technical partners across Risk, Product, Finance, and other areas of the business. Lead the Team and Model Portfolio * Mentor, develop, and grow a team of full-stack Data Scientists and Machine Learning Engineers. * Hire and level talent with the right balance of modeling depth, experimentation expertise, and production engineering capability. * Own the team's delivery and outcomes: what ships, when it ships, how it performs, and whether it holds up in production. * Own the model roadmap across ACH return risk, fraud, and account takeover, including prioritization, model reviews, and postmortems following loss events. * Own model performance in production, including monitoring, drift detection, retraining cadence, model updates, and decisions around when models should be replaced or retired. * Establish a high bar for technical quality, operational rigor, and measurable business impact across the team., * Maintain strong model documentation, validation evidence, and audit trails, and support independent validation of the team's models. * Ensure production decisions are explainable, with appropriate reason codes available when decisions need to be justified to merchants, internal reviewers, or auditors. * Partner closely with Product and Engineering on scoring latency, feature availability at decision time, and integration into Trustly's risk engine. * Co-own decision strategy and the policy layer with Risk Operations, translating model outputs into business outcomes. * Quantify and communicate the tradeoffs between approval rates, customer experience, and loss rates when recommending changes. * Co-own loss forecasting with Finance and Risk, connecting model performance and behavior directly to financial outcomes. * Present model performance, loss drivers, emerging risks, and roadmap tradeoffs to senior and executive stakeholders.