Real-Time Data Tokenization With Cross-System UUID Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing tokenization methods generate non-correlatable tokens across different systems, making it difficult to integrate and use tokenized data in machine learning models, and existing systems struggle with real-time and batch processing of sensitive data while maintaining confidentiality.
Innovation Solution
A system utilizing a key-value database with universally unique identifiers (UUIDs) for tokenization, enabling on-demand tokenization across various computing systems, allowing real-time and batch processing, and ensuring consistent token generation for sensitive data elements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional tokenization methods are used to protect sensitive data, then data confidentiality is improved, but token correlation across different systems deteriorates
Solution Approach 1:
The patent implements a universal tokenization system that generates consistent tokens across multiple different systems and databases. The tokenization service acts as a central authority that maintains a mapping between sensitive data values and their token representations, ensuring that the same sensitive value always produces the same token regardless of which system requests it. This universal approach enables token correlation across systems while maintaining data confidentiality through consistent token generation.
2Productivity
If real-time tokenization is implemented, then data processing speed is improved, but system complexity increases
Solution Approach 1:
The patent introduces a tokenization service as an intermediary component between data sources and data consumers. This service handles the complexity of tokenization logic, mapping maintenance, and token generation in a centralized manner. By placing the complexity in a dedicated intermediary service rather than distributing it across all systems, the patent enables real-time tokenization while managing system complexity through specialized infrastructure.
Solution Approach 2:
The system performs preliminary tokenization of sensitive data before it is used in machine learning workflows or shared across systems. By pre-generating tokens and maintaining mappings in advance, the system enables rapid real-time tokenization operations without requiring complex computations during actual data processing, thus improving processing speed while managing complexity through upfront preparation.
3Reliability
If sensitive data is masked or obfuscated for machine learning training, then data confidentiality is improved, but model training accuracy deteriorates
Solution Approach 1:
The patent creates token copies of sensitive data that preserve the structural and relational properties of the original data while removing identifiable information. Instead of masking or obfuscating the actual values, the system replaces them with token representations that maintain the data's mathematical and logical relationships. This copying approach allows machine learning models to learn from the tokenized data with the same accuracy as they would from the original data, while confidentiality is maintained because the tokens cannot be reverse-engineered to reveal the sensitive values.
Data Source
AI summary
Systems and methods for on-demand tokenization of data. Sensitive data elements are tokenized according to a standardized approach, with the resulting token being, or incorporating, a universally unique identifier (UUID). Each payload-token mapping is stored in a distributed key-value database, with the input payload serving as the basis for the key and the generated token stored as the value. The distributed key-value database is accessible to systems within a trusted environment, and also to authorized external applications via a tokenization API service, allowing a variety of applications to request tokenization in real-time, and their data sets to be cross-correlatable.