TRUSTED BY the LARGEST BANKS, TELecom, AND RETAILERS
One Platform for Launching and Streamlining Data Collaboration with any Parties
Partners’ data is connected using a virtual DWH
Data is accessible via data catalogs and marketplaces, which may be used for data discovery, data modeling, running models with shared data, and activating results in communication channels and processes
A trusted execution environment is used for secure data calculations
A decentralized protocol is used for managing metadata, rights and permissions, scripts approval, billing, and other actions
Build ML Models, get Insights Working with Joint Data with Your Partners Without Sharing the Data or Losing Control Over it
1
Connect Your Data
Register it in the data catalog. Grant permissioned access to your partners
2
Order Partners’ Datasets
Using the Data marketplace
3
Develop and Train ML Models
In a protected datalabs VMs using depersonalized data
4
Automate Calculations
In a Cleanroom with full scale data
5
Activate Results
In communications, processes and analytics
Connect your data
Register it in the data catalog. Grant permissioned access to your partners
Use the platform’s database or your own one (Virtual Data Warehouse)
Work with structured and semi-structured data
Define the necessary distortion methods
Publish and describe your data
Grant access to the specific partners and collaboration cases
Structure Your Partners’ Data
Using the data marketplace
Select the necessary partners’ data and order it
Extract the partners’ data for modeling into protected datalabs
Use direct P2P data exchange if approved by the partners
Develop and Train ML Models
In a protected datalabs VMs using depersonalized data
Use a familiar working environment
The system automatically depersonalizes the data and applies distortion models
Discover your partners’ data, upload additional data, build models
Publish results as scripts, audiences, or reports
Automate Calculations
In CleanRoom with full scale data
Review and approve scripts to be executed on a full scale with your data
Order scripts for scheduled execution or evoke them using API
Control their execution using logs
Avoid any disputes with logs stored at an immutable ledger
Activate Results
In communications, processes, and analytics
Provide joint data insights to your partners with a secure BI
Use API to integrate calculations into your processes
Use the resulting audiences in the communication engine
Create BI analytics or add other features to your database (such as underwriting scores)
Activate audience data on communication platforms
A trusted Environment for Working with Any Parties
Perform calculations with two or more parties
Use decentralized smart contracts to manage access and restrictions
Have full control over all operations using an immutable ledger
Variability of Calculations
Use familiar programming languages: Python, R, PySparc, Scala, Java, and others
Work with structured and semi-structured data
Run any number of simultaneous calculations
Security and Privacy by Design
Hardware based encryption using Intel SGX
Build models with depersonalized datasets in a protected environment
Train models in datalabs or fully confidentially in cleanrooms
Execute models in cleanrooms
Have your personal data protected and invisible
Using Intel SGX has allowed us to build a platform that can securely and privately process data from a variety of partners. This allows us all to collaborate better and create more effective advertising campaigns for our customers. Our team is extremely excited about the upcoming enhancements to Intel SGX that will make the secure enclaves capable of handling even more data, making it possible for us to further scale our data management platform.
The Aggregion solution will allow us to unite efforts in search for the most effective methods of working with a joint client base within a framework of absolute security and confidentiality for both commercially significant information and personal data.
Nikolai Orlov
Beeline Big Data Development Director, VEON Kazakhstan
