.png)
Support for multiple types of models and algorithms, during training and inferencing
APIs automatically scale to handle production workloads.
explainX is also open source and freely available for you to use today.
Uncover the "why" and "how" behind each prediction made by the machine learning algorithm
Identify features that impact the prediction and optimize them.
Generate similar profiles to identify similar behaving clusters.
Debug your machine learning models and identify biases to improve model accuracy.
Explore the relationship between different variables to improve feature selection.
Drill down to monitor model performance in different subsets to address biases.
Identify similar profiles to accelerate compliance and audit
Make customers & business stakeholders happier with responsible and explainable AI solutions
Reduce compliance costs by identifying biases and increasing AI understanding.
Explain overall model behavior and identify biases in your data
Explain the logic of each prediction in real-time
Play around with input feature values to see the potential impact on predictions
Support your analysis by finding similar profiles and counterfactuals
Explore your data, model and rationale behind AI predictions in depth to gain a comprehensive view of your AI model.
Explain your custom models. No matter it is XGBoost, CatBoost, Neural Network or a simple Logistic Regression, we have you covered.
We have built explainX using state-of-the-art techniques including SHAPLEY values, Integrated Gradients, What-If Analysis, in-house developed Contrastive Explanations and Prototypical Analysis
.png)
.png)
