Applied Analytics & ML

Develop robust, resilient, and highly trustworthy predictive analytic and ML models with patented interpretable modeling techniques.

• Get ML-like predictive power and a fully transparent, ethical, compliant approach with segmented scorecards and binning.  
• Perform what-if modeling and simulations on what would happen to your portfolio if you approved previously declined customers with reject inference. 
• Build collaborative and engaging data visualizations, creating deep insights that predict business outcomes.
• Add constrained optimization models to perform in-situ tradeoffs for specified business scenarios.

Manage models across the entire AI lifecycle from development and training through to production.

• Get more analytics into production — faster — with a unified enterprise platform approach to author, test, simulate, and deploy new models. 
• Improve traceability and drastically reduce cost of model development with an enterprise-level decision asset repository, improving collaboration among data science, business, and IT. 
• Leverage FICO’s rich patent portfolio to create explainable-first machine learning model architectures to confidently make real-time decisions with transaction data.

Analyze and link large volumes of federated data inferring and visualizing connections related to identity resolution and social network analysis.

• Build and update relationship networks that derive critical characteristics, improving the power of your predictive models. 
• Proactively detect and prioritize organized fraud and criminal activities using relationship-driven predictive analytics to detect suspicious behavior. 
• Use real-time, configurable search algorithms across enterprise data with a historical overview to find networks of association across data stores. 
• Give investigators an immediate, intuitive dashboard of established matches.

Operationalize both models developed in FICO® Platform and external models from other DSML development platforms with open model execution.

• Mix and match the best open-source and third-party tools, with support for models (PMML, SAS, R, Python, and Scala), notebooks (Jupyter and Zeppelin), libraries (scikit-learn, H2O, xgboost, and Spark ML), and open FICO Platform APIs. 
• Support real-time, streaming transaction data with InstantML for fraud detection, anti-money laundering, open banking, and next best action. 
• Accelerate model development and easily swap data providers with banking characteristic libraries.