Hawk builds AML and financial-crime detection software powered by explainable AI, deployed on Kubernetes + AWS/GCP with Kafka, Elasticsearch, and PyTorch in the stack. The tech mix—heavy on ML infrastructure (Spark, Hadoop, scikit-learn, TensorFlow) paired with real-time streaming—reflects their focus on reducing false positives while scaling detection coverage. Engineering dominance in hiring (9 of 36 roles) and active work on GenAI-powered investigative workflows signal a shift toward interactive compliance tooling rather than static rule engines.
Hawk is a RegTech company founded in 2018, headquartered in Munich, serving financial institutions and fintechs with anti-money laundering, transaction monitoring, payment screening, and fraud detection. The product combines rule-based and machine-learning detection layers to flag suspicious activity while minimizing operator noise. The team spans engineering, product, support, and legal—reflecting the compliance-first nature of the business. They operate across Germany, Malaysia, the US, and UK, with a current workforce of 51–200 people and deliberately slow hiring (4 new roles in the last month).
Hawk runs Kubernetes on AWS/GCP, uses Kafka and Elasticsearch for real-time data, and relies on PyTorch, TensorFlow, scikit-learn, and Apache Spark for ML training and inference. The stack includes Terraform for infrastructure, PostgreSQL for transactional data, and Prometheus/Grafana for observability.
Current projects include GenAI-powered investigative workflows, real-time AI detection, explainable AI methods, and enterprise professional services delivery. Internal priorities also include platform stability, scalability for large customers, and false-positive reduction.
Other companies in the same industry, closest in size