SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Arkose Labs is seeking a Machine Learning Director to lead the ML research and engineering team driving fraud detection and risk-scoring capabilities across the Arkose Titan platform. This is a builder-manager role combining hands-on technical leadership with team management and strategic direction-setting.
You will define and execute the Data Science and Machine Learning roadmap, aligning research initiatives with product priorities and business outcomes. You'll lead the development and deployment of ML models powering real-time fraud and risk decisioning at scale, working closely with detection engineering to translate emerging fraud patterns into features and models.
Key responsibilities include establishing modern MLOps practices—model governance, experimentation frameworks, and end-to-end lifecycle management—while owning model performance in production, including precision/recall tradeoffs, drift monitoring, retraining cadence, and latency/cost optimization. You'll drive applied research in graph machine learning, behavioral analytics, and intelligent fraud detection systems, contributing to patents and publications.
You will manage, coach, and grow a team of ML researchers, handling performance management, career development, and hiring. The role requires staying hands-on with model design, feature engineering, architecture reviews, and code review alongside the team. You'll translate fraud and business metrics into measurable ML objectives and communicate roadmap tradeoffs to leadership and cross-functional stakeholders.
Required: 6+ years building and deploying ML models in production, including 2+ years directly managing ML engineers or data scientists. You need a track record of shipping ML systems that moved real business metrics, ideally in fraud, trust & safety, cybersecurity, or adversarial domains with imbalanced data. Strong technical depth across classification, anomaly detection, graph-based methods, and sequence models is essential. Full ML lifecycle experience (data pipelines, feature engineering, training, evaluation, deployment, monitoring) and proven people-management skills are required.
Nice-to-have: experience with bot detection, device fingerprinting, behavioral biometrics, real-time risk scoring, low-latency ML serving infrastructure, or current ML research in adversarial ML and anomaly detection.
