NOTHING TO SEE HERE
Privately Qualify, Disqualify, & Prepare Datasets for Analysis
TripleBlind’s Blind Data Tools provide data views that enable you to understand and prepare for the use of protected datasets without revealing any sensitive information. These tools give data users the ability to qualify or disqualify datasets, and prepare to operate on the full dataset in a privacy-preserving way.
Find what you’re looking for.
Exploratory Data Analysis
TripleBlind’s web interface allows data scientists to view valuable metadata about any dataset positioned by a data provider. Each field includes summary statistics like mean, minimum, maximum, count, and more.
TripleBlind’s Data Profile EDA report empowers you to determine if the data is relevant for your use case before having to formally negotiate terms of access with the data provider. With TripleBlind’s exploratory data analysis, data owners can provide descriptive statistics and characteristics without ever risking the privacy or security of the underlying dataset
A privacy-preserving preview, displayed just so.
The Asset Overview page for a dataset displays a 10-record set of representative data for the dataset. This Mock Data table is a privacy-preserving representation of the underlying dataset, purposefully prepared by the data provider.
Fields containing sensitive data are masked with random characters or replaced with realistic, reader-friendly AI-generated random values.
For data that is not sensitive, the data provider has the ability to selectively unmask fields considered to be safe for display. The 10-record set of the unmasked field contains a random sample of values from the underlying dataset.
Two or more organizations may have the same data on different samples of a population, but combining that data for analysis is rarely a seamless process. Mock Data allows the data scientist to see the fields, formats, data-types, and example values in disparate datasets without needing to see the raw underlying data, and appropriately account for any differences in their data preparation and algorithms.
Wrangle. Munge. Build.
Blind Sample takes the representative sample shown in Mock Data to the next level. This operation provides data scientists with a realistic privacy-preserving sample similar to the real underlying data.
Request any number of records from Blind Sample, and TripleBlind will use underlying privacy-preserving methods, such as generative adversarial neural networks (GANs), to create sample records that look and feel like the real underlying dataset. The representative sample can be downloaded, examined, and used to develop your process before deploying against the real data.
With Blind Sample, data engineers can:
- view and interact with real data
- refine their understanding
- prototype their processes
They can do this all before implementing the solution against the real, protected (blind) data.
Book A Demo
TripleBlind keeps both data and algorithms in use private and fully computable. To learn more about Blind Learning, or to see it in action, please book a demo!