Model Testing and Debugging
Drive model quality and performance with automated testing, explainability, and root cause analysis
Drive model quality in dev and production.
Comprehensively test your ML for Quality
Root cause analysis (RCA) of issues discovered in production
Get models into production quickly with comprehensive testing and explainability
Most comprehensive techniques for debugging production issues and improving model quality
Robust analytics provide an in-depth look of models and identify areas of improvement quickly.
Robust analytics provide an in-depth look of models and identify areas of improvement quickly.
- Accuracy
- Generalization
- Conceptual Soundness/Reliability
- Bias/Fairness
- Stability
- High Error Segments
- Model Performance
- Data Quality
The first automated Test Harness for evaluating ML models
TruEra’s Test Harness capabilities provide automated, systematic testing for performance, stability, and fairness. Get value out of the box:
Default tests evaluate models against a series of metrics with pre-established thresholds throughout the model development lifecycle. Customized tests are also easily created in the intuitive UI. TruEra minimizes manual test setup time and effort, and eliminates the guesswork for designing tests for new use cases.
Identify top-performing models using a robust set of metrics
Machine learning development is highly iterative and experimental, so quickly understanding the evolution of model versions 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.
Find the causes of unjust bias
Minimizing model bias is not just a good business goal, it is also increasingly required by stakeholders and regulatory bodies. TruEra helps you to:
Minimizing model bias is not just a good business goal, it is also increasingly required by stakeholders and regulatory bodies. TruEra helps you to:
- Identify model bias for protected classes or any identified group
- Hone in on which specific features are contributing most to bias results, for faster explanation and resolution. This deep analysis is unique to TruEra.
Easily deploy and scale
TruEra embeds easily within your existing infrastructure and workflow
- Deploy SaaS or in your cloud, including private cloud, AWS, Google, or Azure
- Work across model development solutions, such as AWS Sagemaker, Microsoft Azure, Vertex.ai, Dataiku, DataRobot, or H2O.ai, and model serving solutions
- Scale to meet high model volumes
- Export data via APIs to Business Intelligence tools such as Tableau and Looker
