Model Life Cycle Engine

Automated Process Workflows

Model Life Cycles are the blueprint for ModelOps

A Model Life Cycle (MLC) automates the requirements and processes for operationalizing a model. It includes detailed process workflows with well-defined steps for operating, governing and maintaining the model throughout its post-development life cycle, until it is retired.

Accelerate model operationalization with pre-defined Model Life Cycles

Use provided steps and processes to automate model registration, operations, risk and monitoring
Customize with your business, risk and technical KPIs and controls
Scale use of MLCs with metadata that defines the model’s unique path and requirements
Unite data scientists, data engineers, IT operations, model operations, risk managers and business unit leaders through clearly defined processes

Automated Registration Processes

Ensure all approvals, validations and documents are complete before models are deployed to production environments or applications. Registration processes include:

Deploy registered models into your QA runtime
Run models through a series of tests
Trigger and automated security scan
Validate required documentation is in production model inventory
Route for appropriate approvals for production deployment

Automated Model Risk Management Processes

Specify and enforce your regulatory and business controls. Track and log all steps and actions for reproducibility and auditability

Maintain comprehensive and updated production model inventory
Validate completion of all required steps, approvals, documentation, background tests, and more
Integrate with your model risk management systems

Automated Operationalization Processes

Accelerate the operationalization of models with automated, repeatable processes that track and orchestrate deployment and change.

Straight through deployment models
Orchestrate and validate change management tasks
Run and create snapshots of model testing and changes
Integrate with IT service management systems

Reflexive Monitoring Processes

Ensure models are performing optimally with automated performance monitoring.

Champion/challenger model comparison
Statistical performance monitoring
Data and concept drift monitoring
Daily Back-testing
Interpretability
Fairness
Model retraining and refresh