Machine Learning Engineer

We are seeking an experienced Machine Learning Engineer specializing in Fraud Analytics to design, build, and deploy scalable models that detect and prevent fraud across our digital platforms. The ideal candidate will have hands-on experience leveraging AWS SageMaker , Bedrock , and advanced ML techniques to build and maintain fraud detection solutions for high-volume transactional systems.

Responsibilities

Model Development & Engineering

Design, develop, and implement machine learning models for fraud detection, prevention, and risk scoring.

Train and deploy ML models using AWS SageMaker (Pipelines, Training, Deployment, Feature Store).

Build and optimize LLM / GenAI solutions on AWS Bedrock to enhance fraud insights (e.g., pattern interpretation, anomaly explanation, synthetic fraud data generation).

Develop and maintain real-time detection pipelines leveraging streaming data (Kinesis, Kafka, or Spark).

Data Engineering & Analysis

Work with large-scale fraud datasets (e.g., credit card transactions, identity risk, e-commerce fraud, account takeover).

Build data pipelines integrating structured and unstructured data from multiple fraud data sources.

Collaborate with data engineering to create fraud features (device fingerprinting, behavioral biometrics, velocity checks, anomaly signatures, rule-based signals).

Fraud Strategy & Collaboration

Partner with risk, product, security, and engineering teams to productionize fraud solutions.

Use ML to enhance rule-based systems and fraud scoring engines.

Evaluate and integrate fraud tools, risk APIs, knowledge graphs, and anomaly detection frameworks.

Monitoring & Compliance

Deploy continuous model monitoring (drift, false positives, recall KPIs).

Ensure compliance with regulatory and audit requirements (PCI-DSS, SOC2, GDPR).

Develop documentation, reporting, and dashboards for fraud model performance.

Required Qualifications

Skill Category

Requirement

Fraud Experience

MUST HAVE at least 2 3 years hands-on experience with fraud datasets and model deployment (e.g., identity fraud, card fraud, transaction risk, synthetic ID detection).

AWS Hands-On Skills

Production experience with AWS SageMaker and AWS Bedrock in ML workloads.

ML Expertise

Experience developing and deploying supervised, unsupervised, and NLP models for fraud (e.g., GNNs, anomaly detection, time-series, ensemble methods).

Programming

Python required. Experience with PySpark, SQL preferred.

Big Data

Experience with Spark, EMR, Kinesis, Redshift, Glue, or Snowflake.

Real-Time Systems

Implemented real-time scoring systems in production.