Senior Data Scientist - Sanctions Screening

We’re looking for a Senior Data Scientist to join our growing Sanctions Screening Team in Tallinn.

This role is a unique opportunity to work behind the scenes of company transactions, understand how we mitigate risk and at the same time provide our customers with the seamless service they deserve. What you build will have a direct impact on Wise’s mission and millions of our customers.

About the Role:

Our screening team is responsible for sanctions, PEPs (Politically Exposed Persons) and Adverse Media screening.

The screening team has a name matching service that is routinely tested against an internal benchmarking suite, and annually against external benchmarking suites.

We are looking for someone to own the testing, tuning and optimisation of these matching algorithms. The candidate will own this function - be responsible for the overall testing and tuning strategy, deep dive testing results, understanding how to optimise rules for efficiency and effectiveness and create rules for complex controls.

Here’s how you’ll be contributing:

• Automatising algorithm testing on real customer data as well as synthetic data. Packaging the service into a library or deploying it to staging or production environments
• Benchmark testing
  • Analyse results from exact name matching and fuzzy name matching against internal and external benchmarks
  • Identify and categorise types of missed cases and map those to known or new issues
  • Propose technical solutions to reduce the number of missed cases identified
• Benchmark creation
  • Design an internal test set to evaluate both the precision and recall of the screening engine
  • Align performance on the test set with real-life production performance
  • Extend internal benchmarking to new scenarios for better coverage of screening algorithms
• Define an overall strategy for testing and tuning
  • Evaluate the use of an internal benchmarking tool
  • Evaluate testing capabilities of external vendors in the market to define the most effective method of continuous external benchmarking, processes of governance
  • Tailor business rules to reduce hit rate, prepare tuning data, reviewing hit reduction strategies and work with product managers, compliance and engineers to ensure roadmap alignment
• Rule optimisation
  • Tweaking engine configuration to find the sweet spot for precision and recall.
  • Provide answers on how many historical true positives we would miss based on different optimisations
  • Work on advanced rules for complicated controls, such as vessel screening
• Uncover and action on opportunities to help the Screening operational team scale
  • Automatising or creating solutions to assist operations in their work
  • Modify existing tooling introducing LLM assistants to improve the efficiency of agents and speed of case resolution

A bit about you:

• Experience implementing, testing and evaluating performance of multiple rules across systems;
• Strong Python knowledge. Ability to read through code, especially Java. Demonstrable experience collaborating with engineering on services;
• Strong algorithmic design and testing skills. A big plus for proven experience with name matching algorithms;
• Experience with statistical analysis and good presentation skills to drive insight into action;
• A strong product mindset with the ability to work independently in a cross-functional and cross-team environment;
• Good communication skills and ability to get the point across to non-technical individuals;
• Strong problem solving skills with the ability to help refine problem statements and figure out how to solve them.

Some extra skills that are great (but not essential):

• Familiarity with automating operational processes through technical solution, for example Large Language Models;
• Knowledge of Sanctions and Name Screening Optimisation and Tuning
• Experience working in a heavily regulated business domain.

#LI-AT1 #LI-Hybrid