Cake is a fully digital bank built on AI as its operating backbone, not a feature layer. The tech stack — Python, TensorFlow, PyTorch, Kubeflow, Vertex AI, plus mobile-first infrastructure (Flutter, Kotlin Multiplatform, Firebase) — reflects a company that has baked machine learning into underwriting, fraud detection, and operations from inception. Current hiring skews heavily senior (10 of 11 open roles) concentrated in engineering and product, paired with active projects around eKYC, soft POS, fraud detection, and credit scoring models, suggesting both technical depth and a push toward ecosystem expansion beyond core banking.
Cake Digital Bank is a fully online bank operating in Vietnam, launched in 2021, serving over 6.5 million customers. The platform handles 100% of transactions digitally with approvals completed in minutes. The bank operates on an AI-native architecture that covers product design, customer acquisition, underwriting, fraud detection, and risk governance. In just over three and a half years, Cake achieved EBITDA profitability as the first digital-only bank in Vietnam to reach that milestone. The product roadmap includes eKYC platforms, soft POS integrations, tap-to-pay capabilities, and real-time fraud detection systems. Operations span Vietnam and the Philippines.
Core: Android, Flutter, Kotlin Multiplatform, Firebase for mobile; Python, TensorFlow, PyTorch, scikit-learn for ML; GCP, AWS, Azure for cloud; Kubernetes, Docker, Istio for infrastructure; PostgreSQL, MongoDB for data; Apache Airflow, Kubeflow, Vertex AI, MLflow for ML pipelines.
Active projects include eKYC platform development, soft POS SDK, tap-to-pay integration, real-time fraud detection, ML model pipelines, credit scoring, and financial services recommendation systems. Also focused on scalable API ecosystems and post-launch product evaluation.
Cake by VPBank - Digital Bank's technology stack, projects, and hiring signals are inferred from public hiring and company data — career pages, public listings, and company web presence — then clustered and de-duplicated. Figures are estimates that refresh over time. Read our full methodology →
This is not an official vendor or customer list. It is a technology-adoption signal inferred from public data, intended for B2B research.