The Technology Behind Model Intelligence
In enterprises, machine learning is used to algorithmically build a predictive analytical model based on training data in order to make predictions and/or decisions as part of a business application. Machine learning is a powerful business intelligence technology but it has a major flaw: It’s a black box.
Model Intelligence helps enterprises dramatically improve the way they build and operationalize models to maximize the responsible adoption of machine learning and AI technology.
AI.Q Enterprise-Class Explainability
The ability to explain each prediction of a machine learning model is core to model intelligence. These explanations help all enterprise stakeholders: data scientists; business, engineering, and operations teams; data subjects; and regulatory and compliance teams.Learn more about AI.Q Explainability
Model Quality Analytics
Model quality helps ensure models achieve the intended business impact. Truera analyzes several facets of model quality including:
- Conceptual soundness
- Segment Analysis
Model Comparisons & Selection
Machine learning development is highly iterative and experimental, so quickly understanding the changes between iterations and across experiments is critical. With Truera, data scientists can more deeply and easily compare models than ever before, enabling them to extract insights to guide faster and more effective model development.
Reporting & Monitoring
It’s hard to maintain black box machine learning model performance and trust over time when new data changes from the training data used to create the model. This is top of mind, especially during the current Coronavirus pandemic. Truera enables data scientists to monitor and understand data drift, concept drift, and model quality over time.
Review & Governance Workflow
High-stakes and regulated models often require separate model validation or governance processes. Truera encodes best practices for validation and governance including:
- Model documentation
Sharing & Collaboration
Truera supports the workflow involved in the full lifecycle of model development, validation, operationalization, and monitoring across multiple model projects and versions. This enables data scientists to easily share, demo, and collaborate on model intelligence insights with their managers, business counterparts, and model validation/compliance teams.
Broad Machine Learning Technology Support
Most data science teams use multiple model development technologies based on team familiarity and platform efficacy for different use cases. Architecturally, model intelligence and governance should be separated from development so that data science teams can establish consistent, best practice-based methodology for evaluating models and pick the best-of-breed model development technologies. Truera enables this architecture by seamlessly integrating with all of the popular model development technologies.