detectus

Detectus is a modular software platform designed for detecting anomalies and events across data streams. It combines machine learning models with rule-based heuristics to flag unusual patterns in real time and in historical analyses. The system is designed to integrate with existing data pipelines, supports streaming and batch processing, and outputs confidence-scored alerts along with explainable rationales.

Detectus was developed by Nebula Systems and released in 2017, with subsequent major updates in 2019 and 2021. It originated to address cross-domain needs such as fraud detection and operational monitoring.

Key features include a modular architecture with plug-in analyzers, support for unsupervised clustering, semi-supervised anomaly detection, supervised classifiers, concept drift detection, and privacy-preserving options such as on-device processing and data minimization. It provides dashboards, alert routing, audit trails, and application programming interfaces for integration.

Industries include finance, e-commerce, healthcare, manufacturing, and cybersecurity. Typical use cases encompass fraud detection, network intrusion, equipment fault detection, and abnormal patient monitoring.

In evaluations, Detectus is praised for scalability and explainability but criticized for reliance on labeled data for some models, potential for false positives, and deployment complexity. Effective use often requires strong data governance and skilled operators.

Related topics include anomaly detection and real-time analytics.