The Account Defender team at HUMAN is responsible for safeguarding user accounts across the internet by identifying and mitigating fraudulent activities post-login, including synthetic account creation, unauthorized account access (ATO), and coordinated malicious behavior. We combine large-scale behavioral data with detection logic to stop sophisticated actors and protect real users.We’re looking for a Senior Data Scientist to join our team and take a key role in shaping detection strategies, researching behavioral signals, and developing models that power HUMAN’s real-time fraud defenses.This is a high-impact role that involves working closely with other researchers, data analysts, product managers, and engineers to build, deploy, and iterate on fraud detection capabilities at scale.
What you'll do:
Lead end-to-end research projects to detect fake users and account takeovers – from exploring behavioral patterns and validating hypotheses to deploying logic in production and evaluating impact.
Design and implement detection logic and features based on large-scale behavioral signals and fraud patterns.
Analyze real-world traffic to surface anomalies and iterate on detection strategies that improve precision and coverage.
Work closely with Product and Engineering to integrate solutions into production systems and influence roadmap decisions.
Help shape the team’s detection methodologies, tooling, and best practices for scalable, real-time fraud mitigation.
Who you are:
You have experience solving large-scale, data-intensive problems in production environments (we process trillions of events per day, so comfort with big data is essential).
You are fluent in Python and SQL, and comfortable with modern data tools and platforms.
You understand the strengths and trade-offs of various statistical and detection approaches, and can ship practical solutions to production.
You are able to bring clarity to ambiguous problems and simplicity to complex ones.
You write maintainable, object-oriented code and have strong debugging skills.
You’ve worked on cross-functional projects with both technical and non-technical stakeholders, and are comfortable following SDLC best practices (version control, unit testing, CI/CD).
Experience in fraud detection, account security, or the cybersecurity space is a strong advantage.
About the team:Account Defender is a research-driven team focused on protecting the integrity of real user accounts. We work with a wide variety of detection tools, including rule engines, velocity checks, clustering techniques, and real-time behavioral signals. Our detection logic is deployed in production and includes both heuristic methods and real-time machine learning models. The team’s work directly impacts customer trust and safety across some of the world’s largest platforms.