Cortex Certifai
Cortex Certifai
By CognitiveScale Inc
Certified enterprise ready
Gain visibility and control over automated decisions by managing the risk and optimizing the value of black-box AI models. Certifai automatically detects and scores model risks including performance, data drift, bias, explainability, and robustness.
Software version
1.3
Runs on
OpenShift 4.6 - 4.9
Delivery method
Operator
Trust is the foundation of digital systems. Without trust, AI cannot deliver on its potential value. However, machine learning models often function inside black-boxes. This has resulted in significant business risks that hampers Enterprise AI adoption. Cortex Certifai detects and scores 6 dimensions of AI business risk: performance, bias/fairness, explainability, robustness, compliance, and data drift.
Automates Data and Model Vulnerability Detection
Certifai takes the guesswork out of understanding risk and vulnerability in AI models by automatically probing the model and testing edge cases. Certifai does not need access to the model's internal code to evaluate it. Cortex Certifai will ensure your AI systems are: - Robust and cannot be fooled - Fair and unbiased toward protected groups - Explainable through clear rationale of AI predictions
Generates Unique AI Trust Index
Certifai generates a numeric score based on key elements of trust. It helps businesses assess the tradeoffs and typical contention between risk and performance. Stakeholders can drill into each evaluation to identify potential improvements. Certifai considers AI risks such as: fairness/bias, robustness, explainability, and key performance metrics like accuracy, precision, or recall.
Provides Reporting for Key Stakeholders
Certifai helps key stakeholders engage in building trusted AI with unique reporting for: - Data science teams - IT experts - Product and marketing executives - Customers and employees - Compliance and risk executives
Pricing summary
Plans starting at
Up to 8 Cores
Automates data, model vulnerability and risk mitigation
Generates Unique AI Trust Index
Specific reporting for all key stakeholders
Additional resources
Want more product information? Explore detailed information about using this product and where to find additional help.
DOCUMENTATION
A Brief Primer on the Certifai App for Data Scientists
Certifai is a risk assessment tool that repeatedly probes a predictive model M in terms of its input-output behavior and provides anevaluation of model risk along 3 dimensions:1 Robustness (R), Explainability (E) and Fairness/Bias (F).
DOCUMENTATION
Why Bias is Interwoven with Several Other Dimensions of Trusted AI
At CognitiveScale, we have grouped the key aspects of building and deploying trustable AI solutions under 5 pillars, representing 5 major types of Risks that Businesses face if these aspects are not properly addressed while employing AI technologies
COMPANY INFORMATION
CognitiveScale Inc
Visit the homepage of the provider.