Overview Join OneSpan as the first ML research hire on the security team to turn telemetry into capable detection and prevention models. You will collaborate with Security to frame data problems, prototype models, and validate them on real data, advancing from research ideas to production-ready solutions. You’ll tackle adversarial, evolving threats in mobile security and shape the ML capability and strategy for the team. This role offers deep technical ownership, close cross-functional collaboration, and meaningful impact on security at scale. Responsabilidades * partner with Security to understand attack techniques and translate into data problems * analyze telemetry from datalake (ClickHouse and S3) to find signals and anomalies * design, prototype, and evaluate attack detection and prevention models (anomaly detection, scarce-label classification, behavioral models, time-series analysis)



* define success metrics and evaluation methodology for ground-truth-scarce problems * progress validated prototypes toward production with engineering and establish datasets, labeling processes * stay current with ML security research (anomaly detection, adversarial ML, fraud) and apply findings Requisitos principales * 5+ years of ML research/development on real-world problems, ideally security/fraud/anti-bot * strong anomaly detection and unsupervised/semi-supervised learning * experience with large-scale telemetry or event data * proficient in Python (pandas/polars, scikit-learn, PyTorch or similar), advanced SQL * ability to work directly against a datalake without prepared datasets * product mindset to deliver evaluable prototypes quickly and communicate results to stakeholders * curiosity about cybersecurity and domain learning * collaboration with cross-functional teams * communication of results to non-technical stakeholders * self-motivation and autonomy * anomaly detection * unsupervised learning * semi-supervised learning

📌 ML Scientist (Barcelona)
🏢 Onespan
📍 Barcelona