privacy enhancing computation tools
Simple integrations bring data and algorithms together for privacy-preserving analysis
The Blind Data and Algorithm API provides a simple-to-use set of libraries enabling users to bring data and algorithms together for collaborative analysis with configurable privacy policies enabled. With the right abstractions, you can leverage the power of privacy-enhancing computation techniques without distracting from your day-to-day.
No expert cryptography knowledge necessary!
How TripleBlind’s Blind Data and Algorithm API Works
Our API is responsible for routing traffic between entities, enforcing permissions, and documenting an audit trail of activity. Just a few lines of code added to the top of your Python or R scripts bring powerful privacy assurances to your data processes.
- Do more with first- and third-party data
- Alleviate stress and headaches for your privacy and compliance teams
Everything you need to know
about TripleBlind in one place.
Connect and Protect
Digital Rights Managed
Digital rights management (DRM) is easier than ever
- Administrators can set fine-grained controls over how their organization’s data and algorithm assets are used.
- Permissions over data use can be managed at the attribute and record level.
- Cryptographic consent is required from both parties (asset users and asset providers) for every operation.
Each party, whether a data provider or algorithm provider, hosts TripleBlind software locally or in their cloud environment. Various collaborating parties discover and connect with one another through the TripleBlind Router. The Router establishes connections and manages the digital rights policies, but steps aside to allow operations to occur in a peer-to-peer fashion.
No Need To Share Datasets or Algorithms
Data Stays Behind your Firewall
TripleBlind removes the need to share datasets or algorithms, which minimizes the risk of a counterparty leaking information or abusing usage agreements. They never hold the keys to access your assets in their raw state. Data users securely run remote calculations, queries, and machine learning model trainings on distributed data through the power of one-way transformations, also known as one-way encryption.
Maintain Data Utility
One-Way Encryption
Most privacy technologies alter datasets to remove sensitive elements so the dataset can be shared. The tradeoff is obvious: the more information you redact, the less utility the dataset has. TripleBlind solves this problem by enabling computation to occur on full, unaltered data while it remains cryptographically obfuscated.
Other encryption methods need to generate a decryption key. TripleBlind’s approach, though, mathematically transforms data for a particular one-time operation, never generating or requiring a decryption key.
This could take the form of a secure multi-party computation (SMPC) on shares of the data, or a federated approach seeded with meaningless noise. Whatever the approach, your computation remains secure and accurate.
Even if the one-way encrypted data were intercepted, it would be irreversibly unintelligible to a malicious actor. And, we offer the same protections for algorithms!
Know what’s being done with your data
Audit Everything
The Blind Data and Algorithm API also allows for accurate cryptographic auditability of every data and algorithm interaction. Not only can admins set controls over data and algorithm usage, they can also view and download audit trails. Administrators have access to audit logging across all their organizations assets and algorithms from a user-friendly web UI.
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!
Blind Compute
Mathematical techniques and privacy primitives used to execute a myriad of computations.
Blind Data & Algorithm API
Our method of securely connecting and managing processes performed with the solution.
Blind Data Tools
Data science tools your teams expect, like pre- processing and EDA, designed around privacy.
Blind Algorithm Tools
Allow easy distribution of models while maintaining full control over your IP.