Unified Data Space

The Lucd Unified Data Space is built as the fused data space with tight coupling into the Lucd AI Library to easily feed TensorFlow models.  The UDS leverages Lucene Search/Indexing and Accumulo NoSQL plus the Data Resiliency of Hadoop to fully secure your data.

  • Easy linkages to select data in the UDS as features to directly feed into an AI model 

  • Role-based access control to each data element enabling controlled rapid and secure discovery, analytics/model building.  Leverages Apache Accumulo is a key/value store based on the design of Google's BigTable

  • Apache Lucene for high-performance, It includes scalable, high-performance Indexing; and powerful, accurate and efficient search algorithms.

  • Server-side programming, stability, and scalability.  

  • A platform with granular access control for each data item ensuring proper data handling while still providing an efficient common repository for data. Secure, compliant, controlled, searchable, fused data for AI models and applications (NIST 800-53, HIPAA, SOX, PCI-DSS, GLBA, CIP, DISA STIG)

The Lucd Unified Data Space Central Repository Capabilities include:
  • Data Fusion any data type needed for Machine Learning

  • Direct linkages to the Lucd AI Library

  • Sizes, shapes, forms data into what is needed for model and provides that via connector into the Lucd AI Library

  • Rapidly ingest new data sources to get immediate value from the data. 

  • The UDS constructs both geospatial and temporal indices

  • The data is self-describing and the UDS is schema-less 

  • Fields can be added, modified or removed as needed without affecting the UDS and without requiring human intervention.

  • Interaction with the data takes many forms, including applications, workflows and most commonly, analytics and visualizations.  Lucd provides a robust set of APIs to interface with the UDS. 

  • Lucd automatically preserves the pedigree of the ingested data upon capture and recording to the UDS.