BlackBox

A distributed environment for storing, managing and utilizing encrypted data sets as well as data processing tools, such as machine learning models, neural networks and ETL pipelines.

Get Started

Why BlackBox?

BlackBox is a distributed environment for storing, managing and utilizing encrypted data sets as well as data processing tools, such as machine learning models, neural networks and ETL pipelines.

BlackBox is...

BlackBox is a unique infrastructural solution providing a layer of trust and self-governance to all the parties of Data Science economy.
BlackBox implements an innovative combination of homomorphic encryption and multiparty computation as well as blockchain-powered asset and governance management as well as the hybrid quality control consensus protocol.

In the current world...

We live in a data-driven world, where data is being generated by every digital process, system, gadget and is augmenting almost any physical object around us. Such an unstoppable explosion of data volume led to a state when data can't be handled manually anymore. As a result demand for efficient processing tools such as machine learning models, neural networks and automated ETL pipelines is constantly increasing.
Data analysis becomes crucial for all the major aspects of decision-making activities, therefore data analysis tools are becoming a key asset in business. As the importance and value of machine learning models rise, IP owners no longer want to publicly share them, though there is a demand for them to be publicly utilized.
On the other end, as data becomes valuable on its own, data owners don't want to expose it even though they desire to use analytical tools for it.
That is where BlackBox comes into play.

Public while private...

BlackBox allows to storage, management and utilization of intelectual assets (eboth data and data processing tools) publicly while still secure, keeping actual "property" private. Such approach creates new data science business case of "renting" intelectual property (IP) instead of selling it.

Use Cases

AgriFood, Healthcare and Financial sectors more and more rely on data. Though there are few drawbacks for each side of the data relationship.

...

For Data Owners

Data owners/creators are very sensitive regarding their data, though they usually don't have the expertise or resources needed to process it. With BlackBox Data owners can store data in an encrypted and secure way while having full control of who, when and how accesses it.


...

For AI Developers

On the other end, data analysts value their creations and would like to commercialise them to multiple clients without given away intellectual rights. BlackBox encrypts data analysis tools and enables owners to manage them securely over API or WEB interface. The unique combination of HE and MPC ensures that models can't be compromised or stolen, while still being suitable for processing and utilisation. HE - encrypts the source code and weights, MPC - ensures each of few computing power providers accesses only part of the model, eliminating risks of reverse engineering attacks.


...

For AI Developers

What is more, data analysis tools can't be created without training data in the first place, and only a few of the stakeholders are ready to provide it out of charge, which slows model development dramatically. With the functionality provided by BlackBox, developers can "rent" data sets and create solutions based on real market data.


...

For Businesses

Multilevel hybrid privacy provision enables independent security of each and all parties involved in the Data Analysis activity. None of the involved parties (Data Owner, Model Owner, Computing Power provider, Customer) can intentionally or unintentionally access the intellectual property during the process as well as gain unauthorised access to the results of such process.


Solution

Explorer

Launch explorer

Contact us to get more information!

Address


Mokslininku st. 2A, Vilnius, Lithuania

Phone


+370 630 78989