After more than a decade of research at Stanford, DARPA and NSF, Ayasdi has created an extremely powerful, scalable solution that allows for analysis of large quantities of complex data.
Ayasdi Iris is query-free, model-free, and code-free, and runs through hundreds of algorithms to automatically construct the underlying shape of the data, revealing hidden insights and anomalies for further investigation.
The Ayasdi Platform has an open API allowing third party tools, data services, and organizations to extend, leverage, and enable ease of integration with their services.
The Ayasdi Platform uses TDA to discover distinct patterns or anomalies within data. Displayed as topological networks, TDA provides high-dimensional analysis of entire data sets, including any proprietary and public data used.
The Ayasdi Platform automatically uses many different machine learning algorithms, including unsupervised, supervised, semi-supervised learning and statistical tests. These are tied together using TDA.
The Ayasdi Platform leverages state of the art distributed cloud computing. Running on top of Apache HBase, computation is distributed over a computational substrate using a proprietary distributed system.
Cancer tumors characterized by somatic mutations colored by clinical outcomes
Identified cancer subtypes that respond differently to treatments
Development of targeted cancer treatments
Genetic variants mapped to individuals and colored by corresponding diseases
Unprecedented discovery of genetic variants corresponding to multiple diseases
Leverage genetic information for personalized medicine
Breast cancer patients displayed by their gene expression levels
Discovered new patient survival subgroups
Development of personalized treatments
Map of 1,092 individuals based on 250,000 genetic variants
Understand global, regional and individual genetic patterns of the human population simultaneously
Characterization of genomic variants that distinguish ethnicities
NBA players displayed by performance data, 2010-11 season
Discovered players group into 13 role categories
Definition of strategies for talent acquisition
Congress members displayed by their voting behavior over the last 20 years
Identified voting patterns within parties by political issues
Public understanding of U.S. Representative voting tendencies
