Senior ML Data Scientist – Fraud & Risk

Triwill Group; New York, Northern (NY, KY); Hybrid; USD 167,000 - 251,000; Full time

Job summary

Mercury is seeking a Machine Learning-focused Data Scientist to strengthen our Risk team and protect customers from financial crime. You will build, validate, and deploy real-time ML models to identify fraud, ensure data quality across pipelines, and collaborate with Risk Strategy and Engineering to optimize deployment and observability.

Ideal candidates have 5+ years of data experience with 3+ years in ML, proficient SQL and Python, and proven production deployment of models in fast-paced

Qualifications

• 5+ years of experience working with and analyzing large datasets to solve problems and drive impact, with 3+ years of ML experience.
• Proficiency in SQL and experience using it to understand and manage imperfect data.
• Proficiency in Python and experience with statistical modeling and machine learning.
• Experience deploying and monitoring machine learning models in production.
• Comfort working in a fast-paced environment with evolving priorities.

Responsibilities

• Build, validate, and deploy machine learning models to identify and prevent fraud in real time
• Support the reproducibility and robustness of said models through documentation, testing, and monitoring
• Ensure data quality and reliability across pipelines and tools
• Collaborate with Risk Strategy to ideate on model inputs and applications and with Engineering optimize deployment and observability

Skills

SQL

Python

Machine Learning

Model Deployment

Data Analysis

Tools

dbt

Job description

Mercury is seeking a Machine Learning-focused Data Scientist to strengthen our Risk team and protect customers from financial crime. You will build, validate, and deploy real-time ML models to identify fraud, ensure data quality across pipelines, and collaborate with Risk Strategy and Engineering to optimize deployment and observability.

Ideal candidates have 5+ years of data experience with 3+ years in ML, proficient SQL and Python, and proven production deployment of models in fast-paced