Datasets underpin every operation in AI and ML. Datasets are what fuel today's AI/ML operations. Most 'datasets' are really just 'data sources'. A true dataset is based on structured information, rows and columns, better known as a matrix as opposed to unstructured data sources. Datasets are the new attack surface in AI/ML systems. If a threat actor changes a value in a dataset, it could result in loss of human life within a pharmaceutical development pipeline or loss of billions for an quantitative asset management company. Vector Space Biosciences provides dataset provenance and security. Immutable dataset provenance and security are of paramount importance to all companies operating mission critical AI/ML applications. The best way to provide dataset provenance and security beyond standard revision control systems (which can be compromised based on system clocks and other vulnerabilities is based on hashing datasets on the Ethereum blockchain.

Secondarily, the Dataset Provenance & Security (DPS) pipeline enables enterprises and institutional customers to access tiered service levels based on the VXV utility token and VXV wallet-enabled API key as described in the architectural diagram below. An understanding of API key management along with what it takes to SHA2 hash a matrix made up of rows and columns and then re-hash with a wallet address including other metadata, is a critical factor in the overall architecture.