Machine Learning Engineer, Fraud & Abuse Detection
Core
Build and operate real-time fraud and abuse detection systems for outbound messaging, stopping phishing, smishing, brand impersonation, and artificially inflated traffic.
Role type
Senior IC machine learning engineer (fraud detection)
Builds
Inline hot-path decisioning systems, decision audit ledger, and entity-level risk state for the messaging platform
Domain
Telecommunications / Trust & Safety
Deliverable
production ML models
Required skills
async Python webservice frameworks, text classification, LLMs as classifiers/judges, production ML deployment, rigorous evaluation methodology, adversarial ML
Preferred skills
A2P messaging abuse knowledge, high-scale data systems (ClickHouse, PostgreSQL, Redis, Kafka), cloud-native stack (Docker, Kubernetes, Prometheus), distributed batch scoring (Ray)
Responsibilities
Design, train, evaluate, and operate text classifiers end-to-end; Run shadow-mode evaluations; Build detection against adversarial content; Serve models inline under latency budgets; Power human-in-the-loop fraud review; Collaborate with cross-functional teams
Seniority
Senior, hands-on IC
