Machine Learning Engineer - Fraud & Identity Security

About the role

Role Overview

Ziverge is an engineer-led software consultancy that helps organizations build

high-reliability systems using JVM technologies, cloud-native architectures, and proven

engineering practices. We provide elite development teams, custom software solutions,

and technology consulting for customers who care deeply about correctness,

performance, and long-term maintainability.

We are looking for a Machine Learning Engineer (Fraud & Identity Security) with core

skills in Python, Apache Spark, and SQL to join our professional services team and

work on long-term, production-critical machine learning systems for identity security and

fraud prevention. You'll build and deploy models that detect and prevent credential

stuffing, account takeover (ATO), and other fraud behaviors, with significant autonomy

and end-to-end ownership.

This is a full-time, remote position. This role requires candidates based in the US or

Canada, with strong working-hours overlap with the rest of the data team.

Key Responsibilities

Understand business objectives and develop models that help identify and prevent credential stuffing, account takeover (ATO), and other fraud behaviors, along with metrics to track progress.

Explore and visualize data to understand the problem space, identify differences in data distribution, and select suitable ML algorithms.

Apply machine learning, statistics, and data mining to improve efficiency across every aspect of identity security.

Develop scalable and efficient methods for large-scale data analysis and model development.

Define data augmentation pipelines, train models and tune hyperparameters, deploy models to production, and monitor and evaluate ML model performance.

Investigate and resolve production issues, contributing to the ongoing reliability of deployed models.

Collaborate with developers, program managers, and product managers in an open, creative environment.

Analyze feature requirements, assess technical feasibility, and provide clear estimates and risk assessments.

Write technical proposals and architectural documentation for new features and system changes.

Plan and implement work across epics and user stories, from initial design through to deployment and support.

Participate in code reviews, share knowledge with teammates, and contribute to a culture of continuous learning and improvement.

Requirements & Skills

Bachelor's, MS, or PhD in Computer Science, EE, or another quantitative discipline.

Minimum 3 years of experience in large-scale machine learning, user behavior analysis, and fraud detection at leading internet companies; experience in the security domain is highly preferred.

Core stack: Python, Apache Spark, and SQL. (We don't need a pure functional-programming engineer for this role.)

Proficiency with machine learning libraries and frameworks: scikit-learn, pandas, and TensorFlow/Keras or PyTorch.

Expertise in visualizing and manipulating large-scale datasets.

Ability to own deliverables end-to-end: requirements, design, implementation, testing, deployment, monitoring, and operational support.

Strong problem-solving skills and the ability to work independently with a high degree of ownership and accountability.

Comfortable working in an Agile environment, collaborating with distributed teams; excellent written and oral communication skills.

Passion for technology, openness to interdisciplinary work, and experience building data-driven services and applications.

Based in the US or Canada, with reliable internet connectivity and strong working-hours overlap with the rest of the team.

Nice to Have

Familiarity with Scala and data engineering — not required, but a plus.

Familiarity with Java/Scala more broadly.

Prior experience working in a consulting or professional services setting, delivering services to external clients.

About the Ideal Candidate

The ideal candidate is a strong data scientist / machine learning engineer with a passion

for applying ML to real-world identity security and fraud problems — someone who

enjoys owning the full model lifecycle, from data exploration through production

deployment and monitoring, in an open and collaborative environment.