Concept: Storage of raw data in its native format.
Various raw data sources — documents, databases, logs — flow into a central "lake" storage area, which then feeds into multiple analysis tools. Unlike a data warehouse, there is no prior transformation; data is kept in its original state until needed.
Concept: Focused subset of a data warehouse for a specific department.
A smaller, specialized subset of data extracted from a larger source and directed toward a specific business unit (finance, marketing, sales). Enables fast, targeted analytics without the overhead of the full data estate.
Concept: Decentralized socio-technical approach to data management.
Different domains manage their own data pipelines independently. Each business domain owns its data products end-to-end — ingestion, transformation, quality, and access. Scales well in large organizations by avoiding a centralized bottleneck.
Concept: Interconnected network of hardware, software, and data.
A circular architecture with a central hub and interconnecting arrows flowing to and from different platforms and tools. Represents the full landscape of data infrastructure — from ingestion to analytics and everything in between.
Concept: Central repository of integrated data from one or more disparate sources.
Structured data flows through an ETL (Extract, Transform, Load) process into a central, organized repository. Optimized for analytical queries and reporting rather than real-time transactional operations.
Concept: Series of data processing steps from source to destination.
A linear flow showing data moving from a source system, through one or more processing stages (cleaning, transformation, enrichment), to a destination — whether that's a warehouse, lake, or analytical tool. Foundation for automated data workflows.
Concept: Unified architecture providing a consistent view of data across the organization.
A layered architecture that "weaves" together different data sources across hybrid and multi-cloud environments. Provides metadata-driven, automated data integration — reducing manual data engineering work while ensuring data is discoverable and governed.