Zest Automated Machine Learning empowers financial institutions with the tools to build, document, and monitor machine learning credit underwriting models in production. These tools allow financial institutions to be compliant with federal regulations, including FCRA, ECOA, and SR-117. Machine learning models provide proven results. ZAML customers, on average, experience a 15% increase in approvals while holding risk constant. Similarly, ZAML customers, on average, experience a 30% decrease in losses while holding originations constant. Depending upon the size of the company, these results can deliver millions to billions of dollars in profit.
ZAML targets the financial services vertical as well as specialty auto finance and insurance verticals. The business audience within these institutions are the Chief Credit Officer, the Chief Risk Officer, and the head of lending for a business unit (auto loans, credit cards, etc.).
ZAML is a suite of tools. These tools include:
ZAML Model - allows users to perform exploratory data analysis and feature engineering to build optimal models.
ZAML Fair - allows users to maximize model fairness by automatically identifying next best alternatives.
ZAML Analyze - enables users to understand a model's impact on a company's economics as well as to perform champion - challenger analysis.
ZAML Deploy - packages real-time explainability for an array of hosting platforms.
ZAML Monitor - empowers users with the ability to monitor the model, including input, output, and system monitoring
ZAML Explain - identifies feature importance, disparate impact analysis, and generates adverse action key factors.
ZAML AutoDoc - provides automated documentation of ZAML models, including model risk management and other model risk reports.