Multi-Tenant Blockchain Metadata for Scalable AI Decision Auditing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Distributed Ledger Technology (DLT) and blockchain platforms face inefficiencies due to fixed, immutable data storage lacking context and metadata, leading to wasted storage space and computational resources, and non-standardized data formats causing issues with data transferability and scalability, particularly in managing complex transactions and high-frequency updates.
Innovation Solution
Implementing a multi-tenant blockchain platform using Distributed Ledger Technology (DLT) in conjunction with a cloud-based computing environment, which includes a blockchain interface for multiple tenants, metadata definition management, and smart contract engines to enable dynamic metadata validation and efficient data storage and retrieval, allowing for standardized data formats and improved scalability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If fixed, immutable data storage is used in blockchain, then data integrity and security are improved, but storage efficiency deteriorates due to wasted space and lack of context
Solution Approach 1:
The patent segments data storage into two parts: immutable hash references stored on-chain for integrity verification, and actual data with metadata stored off-chain or in structured formats. This allows maintaining data integrity through cryptographic hashing while improving storage efficiency by not duplicating entire data sets on every blockchain node.
Solution Approach 2:
The patent introduces metadata dimensions (context, data type, format information) to the traditional blockchain storage model. By adding these dimensional layers of information structure, the system achieves both immutability of core data references and efficient organization/retrieval through metadata indexing without compromising security.
2Adaptability or versatility
If standardized data formats are implemented, then data transferability and scalability are improved, but device complexity increases due to metadata management requirements
Solution Approach 1:
The patent creates a universal metadata schema that serves multiple functions: data validation, format specification, indexing, and transferability enablement. This single metadata structure handles diverse data types (strings, numbers, arrays, objects) through standardized schemas, reducing the need for separate management systems for each data type while improving adaptability.
Solution Approach 2:
The patent uses parameter-based metadata schemas where data structures are defined by configurable parameters (data types, constraints, formats). This allows the system to adapt to different data requirements through parameter configuration rather than structural changes, reducing complexity while maintaining versatility across different use cases.
3Adaptability or versatility
If complex transactions and high-frequency updates are supported, then functionality is improved, but computational overhead increases
Solution Approach 1:
The patent extracts computationally intensive operations (data validation, format verification, complex logic execution) from the main blockchain consensus process and moves them to pre-processing stages or off-chain computation. Only essential state changes and hash references are committed to the blockchain, reducing the computational overhead per transaction while supporting complex transaction types.
Solution Approach 2:
The patent performs preliminary validation, formatting, and metadata assignment of data before it enters the blockchain transaction pipeline. By preparing data in advance with proper schemas and constraints applied, the actual blockchain processing requires minimal computational overhead, enabling high-frequency updates while supporting complex transaction logic.
Data Source
AI summary
Exemplary systems, implement a multi-tenant blockchain platform for managing the Einstein cloud platform's decisions using Distributed Ledger Technology (DLT) in conjunction with a cloud based computing environment. The system operates a blockchain interface to a blockchain on behalf of a plurality of tenants of the host organization, configures the blockchain to share a training data set between two or more tenants pursuant to a consent agreement, trains an AI model to make recommendations based on the training data set, receives a request to register the AI model with an audit record keeping service, receives a transaction at the blockchain, issues a decision by the AI model to accept or reject the transaction; and then proceeds to transact a new asset onto the blockchain recording the decision to accept or reject the transaction and the data set utilized to train the AI model with a version of the AI model.