Ayasdi Unveils Framework to Simplify and Speed Development of Intelligent Applications across the Enterprise
Ayasdi Model Accelerator™ saves valuable time and resources by automating feature selection to create bias-free, highly defensible models in a collaborative environment
MENLO PARK, CA – June 21, 2017 – Ayasdi, a leader in machine intelligence software and a pioneer in enterprise-class intelligent applications, today announced the immediate availability of a new intelligent application called Ayasdi Model Accelerator. Focused on the financial services industry, it is particularly well suited for complex, high-dimensional, modeling problems where the principal goal is to reduce the number of features and variables being used for modeling purposes. The software facilitates the process of building, documenting and deploying those models into production while creating exceptional transparency for regulators and internal model-review boards.
Ayasdi Model Accelerator is the latest addition to the company’s expanding suite of intelligent applications. These applications address mission-critical business problems in financial services, healthcare and the public sector. They package powerful intelligence with intuitive user-interfaces so the broadest audience within an enterprise can consume and act on it to benefit the organization.
Built on Ayasdi’s enterprise AI platform, the Model Accelerator application excels at a variety of modeling challenges but is particularly effective at risk-modeling problems such as loss-given default (LGD), probability of default (PD) and other regulatory modeling problems. These models typically make use of simple regression techniques to support the simplicity and explainability requirements associated for regulatory oversight; however, Ayasdi Model Accelerator supports a range of approaches and it can model exceptionally complex problems.
“With Ayasdi Model Accelerator, enterprises can dramatically reduce the time and resources it takes to create, validate, document and deploy models,” said Ronaldo Ama, EVP, Product and Engineering at Ayasdi. “This is enabled by its carefully designed application workflow and collaboration features that let model designers actively engage with line-of-business owners and internal review managers throughout the construction and documentation process.”
In addition to its ability to rapidly and collaboratively construct and deploy models, Ayasdi Model Accelerator transforms key elements of the model lifecycle. During the model-creation process, the intelligent application documents both variable selection and modeling methodology, systematically and deterministically. It produces a repeatable process with consistent supporting reports that document model lineage, variable selection and cross validation.
With Ayasdi Model Accelerator, not only are the initial variable selections recorded and documented, the entire modeling and approval process is tracked and cataloged for all model developers and collaborators, which facilitates everything from review to model re-use.
These features of the Model Accelerator application and its ability to document and justify recommendations, are a key element in Ayasdi’s framework for all its intelligent applications:
- Discover: via unsupervised (and semi-supervised) learning from data without requiring people to ask the right questions. It discovers relationships on its own.
- Predict: Once data is understood, Ayasdi adds supervised approaches (e.g. classification, regression, ranking) to predict what will happen in the future.
- Justify: Ayasdi adds a critical element of “explainability” to back up its prediction engine, so assertions can be validated, transparent and trusted.
- Act: Intelligent applications are designed to detect and react when data distributions evolve. They need to be “on the wire” to detect and help organizations react to phenomena before it becomes a problem.
- Learn: Intelligent applications need to detect and react as data evolves to be continuously learning. They learn, live in the workflow and are constantly improving. Anything less is simple analytics.
For more information on Ayasdi’s intelligent applications, visit: https://www.ayasdi.com/solutions.
To request a demo of Ayasdi, visit: https://www.ayasdi.com/request-a-demo.
Ayasdi is on a mission to extract value from the world’s data through the creation and deployment of enterprise-class artificial intelligence. Developed by Stanford computational mathematicians, Ayasdi’s unique approach combines machine learning, abundant compute and user-centric design to create intelligent applications that fundamentally change how organizations tackle their most complex challenges.
Based in Menlo Park, CA, Ayasdi is backed by Kleiner Perkins Caufield & Byers, IVP, Khosla, CenterView Partners, Draper Nexus, Citi Ventures, GE Ventures and Floodgate Capital.
For more information, visit: www.ayasdi.com.