Mobile network-backed identity verification and fraud prevention for Canadian financial services
EnStream operates as a trilateral joint venture of Bell, Rogers, and TELUS, delivering identity verification and fraud detection services to banks, payments, gaming, and logistics companies across Canada. The tech stack—Python, PySpark, scikit-learn, PyTorch, TensorFlow, and AWS—reflects a machine-learning-heavy organization building trust and fraud-scoring models. Active project work on identity trust scoring, unsupervised learning, model drift detection, and fraud operationalization signals a shift from rule-based verification toward real-time ML-driven risk assessment. Leadership-weighted hiring (3 leads, 1 director across 8 open roles) suggests capacity gaps in execution and team scaling.
EnStream is a private joint venture of three major Canadian mobile carriers, providing real-time identity verification and authentication services backed by direct access to telecom network and account data covering over 90% of Canadian consumers. The company offers name, address, date-of-birth, and account status verification, plus ancillary services including device authentication, location services, and name-address matching analytics. End-user consent is required for all third-party data sharing. Customers span banking and AML compliance, retail credit issuance, gaming and lotteries (location verification), roadside assistance (vehicle tracking), and transportation logistics. The company operates from Toronto with an 11–50 person team.
Python, NumPy, pandas, PySpark, scikit-learn, PyTorch, TensorFlow, matplotlib, Seaborn, Ray Tune, Optuna, PyTorch Geometric, NetworkX, SQL, Java, and AWS. Heavy emphasis on ML libraries and distributed computing.
Identity trust and fraud-prevention models, unsupervised/semi-supervised learning, model monitoring and drift detection, customer success platforms, customer health scoring, fraud-scoring operationalization, and test automation for distributed services.
Other companies in the same industry, closest in size