The ORIGIEN Decision Engine System is a bank-wide online risk decision system based on big data technology, providing unified management of the full lifecycle of models for various business scenarios.
The ORIGIEN Decision Engine System is a bank-wide online risk decision system based on big data technology. It uses a microservices architecture, supporting multi-threading and high concurrency, with modular and visualized variable configuration. It is designed to meet decision-making needs in intelligent risk control, anti-fraud, marketing, and other scenarios. The system is horizontally scalable, supports customized configuration, and provides unified management of the full lifecycle of models for different business scenarios.
- Data Source Access
Provides visual configuration access for third-party external data sources, supports access to various types of database data, real-time message data sources, and file data source uploads and unified management.
- Variable Preparation
Provides configuration and unified management of input variables, basic variables, and derived variables, with support for visual configuration of processing logic for derived variables using expressions.
- Model Configuration
Provides visual configuration for various types of strategy models, including decision flows, rule sets, scorecards, decision tables, decision trees, decision scripts, and machine learning models.
- Model Testing
Provides batch data testing and historical backtesting for models, with support for viewing historical test records and downloading test results.
- Model Publishing
Supports model publishing as a service, providing decision model calling services of stream-batch integration for upstream business systems.
- Model Execution
Provides model invocation configuration and supports viewing execution details for each invocation and decision report content.
- Full Lifecycle Management
Supports full lifecycle management of models from data access, variable preparation, model configuration, model testing, model services, and execution details.
- Visual Model Configuration
Combines various types of visual model configurations, including rule sets, scorecards, decision flows, decision trees, decision tables, and decision scripts, without the need for code processing.
- Stream-Batch Integration
Supports decision model services of stream-batch integration, including Python, scorecard-type model execution and management of monitoring alerts.
- High Security and High Stability
Ensures continuous, efficient, and reliable 7x24 operation, with a comprehensive disaster recovery plan for high availability and automatic traffic switching in case of failures, ensuring system stability and availability.
- Multiple Deployment Options
Supports various deployment solutions such as load balancing mode, containerization, distributed, and centralized deployment, compatible with common systems like CentOS, Redhat, and Suse.
