Introduction
About The Role Better models lead to better decisions, stronger controls, and safer financial systems. That's where you come in.
As a Data Scientist, you'll own model development for defined financial crime use cases, from understanding client requirements and preparing data to building, evaluating, and deploying models into client workflows. You'll work on challenges such as imbalanced alerts, model drift, and explainability while partnering with client compliance, risk, and technology teams.
You'll also apply NLP, classification, and LLM-based approaches where they improve outcomes, and communicate results through workshops, pilots, and client presentations. This is a hands on role where you'll own assigned use cases end to end, exercise sound judgment on model performance and risk, and help junior data scientists grow.
Job Description
Responsibilities - Own model development. Build, evaluate, and deploy models for assigned financial crime use cases, taking work from requirements through testing and client delivery.
- Turn data into usable models. Analyze large datasets, prepare features, and apply sound statistical and machine learning methods to solve defined business problems.
- Solve difficult modeling problems. Address imbalanced alerts, model drift, explainability, and performance issues through rigorous evaluation and judgment.
- Apply AI where it adds value. Build NLP and classification solutions, including LLM based text analysis, while validating outputs for accuracy, explainability, and business relevance.
- Make models understandable. Use explainability methods such as SHAP and LIME and clearly present model behavior, results, and limitations.
- Partner with clients and teams. Work with client compliance, risk, and technology teams, as well as Data Engineering and delivery teams, on requirements, workshops, pilots, and production integration.
- Deliver into client workflows. Support model deployment, monitoring, and integration so pilots can move into client production successfully.
- Help others grow. Mentor junior data scientists and share practical modeling, evaluation, and delivery practices.
You may be a good fit if you have - 3 to 7 years of data science or machine learning experience, including 2 or more years in financial services.
- Strong model building and evaluation skills for financial services use cases, with a solid grounding in statistics and machine learning.
- Strong Python skills with practical experience using Pandas, NumPy, scikit-learn, or similar tools in production quality work.
- SQL and database experience with Oracle, DB2, SQL Server, Postgres, or similar platforms.
- Explainability experience using methods such as SHAP and LIME to interpret and communicate model outputs.
- Financial crime domain knowledge across AML, fraud, compliance workflows, or related areas.
- Client-facing communication skills. You can lead workshops, present model results, and explain technical findings to compliance, risk, and technology stakeholders.
- Delivery experience. You have delivered at least one model into a client workflow and can own a use case from requirements through pilot delivery.
- An ownership mindset. You know when to resolve modeling challenges independently and when to escalate scope, architecture, or model risk decisions.
Strong candidates may also have - Experience with NLP, classification models, or LLM based text analysis.
- Experience with vector databases and retrieval components.
- Knowledge of AML, KYC, FATCA, or CRS.
- Experience with Spark, orchestration tools, Kubernetes, Kubeflow, Dask, or RAPIDS.
- Experience supporting model deployment, monitoring, and integration with data pipelines.
- A master's degree in a quantitative field such as computer science, mathematics, statistics, engineering, or physics.
- Experience mentoring junior data scientists.
- Curiosity about how AI can improve financial crime detection while maintaining explainability and responsible human oversight.
#LI-Hybrid
