DataDNA: State of AIOps & Observability
Year: 2025
Type: preprint
Abstract: The Foundation: Data DNA TriadsThe core principle behind HGS' Data DNA is the structured pairing of three fundamental datacomponents: X (Taxonomy): The complete categorization of enterprise data. Y (Ontology): Established relationships that define how data entities interact. Z (Knowledge Graphs): Conditional, context-driven associations that evolve over time.At any given point, X, Y, and Z can take on multiple values or vectors. However, their combination ata precise moment forms a unique Data DNA sequence that distinctly identifies an enterpriseprocess, data access point, or procedural interaction.For instance, a sample data operation may be represented as: {X1, Y2, Z3} fetched from {X2, Y1, Z4}for {X5, Y1, Z3}. This structured tracking enables enterprises to monitor lineage, detect anomalies,and ensure data integrity with precision.HGS' Data DNA framework is a transformative approach that redefines enterprise data observability.By leveraging taxonomy, ontology, and knowledge graphs, this model establishes a robust system fordata lineage, security, and operational intelligence. The integration of AIOps, HILF, and RLHFmechanisms ensures continuous learning and adaptation, positioning enterprises to proactivelymanage risks and optimize data-driven operations in real time.This pioneering methodology empowers businesses with intelligent, self-learning systems that notonly enhance data security and compliance but also unlock new opportunities for innovation inenterprise data ecosystems.
Author: Tadepalli, Sharath
Institutions: Marquette University, Purdue University West Lafayette
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7081571031
Cites:
Cited by:
Related to: