Senior AI / Machine Learning Engineer — Fraud Detection

Adobe•Washington, DC

1d•$151,800 - $265,350•Hybrid

About The Position

We are in search of an experienced Senior AI/ML Engineer to develop and broaden fraud and abuse detection systems. You will improve ML techniques in anomaly detection, user/device risk analysis, identity and service abuse, and evolving AI abuse scenarios. This is a hands-on role spanning ML, data, and backend systems, with opportunities to apply LLMs, AI agents, and modern ML techniques to strengthen detection. You will own solutions end to end — from signals and modeling through production deployment and real-time decisioning. Join us in crafting best in class fraud detection systems that will make a significant impact!

Requirements

8+ years building and operating production ML systems, ideally in fraud, abuse, risk, identity, trust & safety, or other adversarial domains.
Solid ML background with practical experience in Python, SQL, and current ML frameworks like PyTorch.
Experience guiding ML systems from feature engineering to production deployment and monitoring.
Strong software/data engineering skills across ML, backend, and data infrastructure.
Experience building with LLMs and/or AI agents, particularly for AI/generation-abuse use cases.
Strong technical judgment, ownership, and ability to solve ambiguous, adversarial problems.
Bachelor's or equivalent experience in Computer Science, Statistics, Mathematics, or related field; advanced degree a plus.

Nice To Haves

Device fingerprinting, identity verification, behavioral signals, network intelligence, or VPN/proxy detection.
Real-time risk evaluation and automated control systems.
Human-in-the-loop or AI-assisted evaluation systems.
Distributed systems and high-scale data pipelines.
Strong adversarial approach — anticipating how attackers adapt to mitigations.

Responsibilities

Build and deploy high-precision ML models for fraud and abuse detection, anomaly detection, and risk scoring.
Engineer risk signals from large-scale account, device, network, behavioral, velocity, and session data.
Integrate ML/AI features into real-time risk decisioning and automated enforcement systems.
Apply LLMs and AI agents to expand detection, investigation, and classification capabilities.
Translate emerging attack patterns and relevant research into new models, signals, and mitigations.
Evaluate solutions across accuracy, latency, cost, and customer impact.
Own model evaluation, monitoring, and drift as attacker behavior evolves.
Partner across engineering, product, and risk teams to ship production-ready capabilities.