Data, Business, Application, and Technical Architecture: Clear Differences Explained

Architecture is a structural description of a system, indicating its components and how they collaborate. Different perspectives create four architecture types—business, data, application, and technical—collectively known as the 4A framework, core to enterprise digital transformation.

Overall Relationship

The four architectures are not isolated; they follow a strict chain: Business Architecture → Data Architecture → Application Architecture → Technical Architecture.

Business Architecture

Business architecture is the starting point, defining what the enterprise does, the capabilities needed, and the relationships between them. A common mistake is focusing on technology before clarifying business needs.

Core Elements

Value Stream : the end‑to‑end process from customer contact to value delivery (e.g., market acquisition → contract → procurement → production → delivery → after‑sale).

Business Capabilities : stable abilities such as supply‑chain management, finance, and HR, usually organized in 2‑4 levels.

Business Processes : detailed steps for each capability, covering objects, activities, rules, and roles, often refined to 5‑7 levels.

Method: first clarify strategic goals, identify business scenarios, break them into capabilities, then detail each capability’s processes.

Data Architecture

Data architecture bridges business and IT layers, translating business objects into standardized data specifications.

Main Flow

Data Subject Area → Conceptual Model → Logical Model → Physical Model.

Data Subject Area : partitions based on business value chain (HR, finance, supply chain, marketing) and identifies core data objects.

Conceptual Model : defines relationships among core business objects, similar to an ER diagram, without field details.

Logical Model : refines the conceptual model to table granularity, specifying fields and relationships.

Physical Model : maps the logical model to concrete database implementations, including types, constraints, and indexes.

Design principles include data‑application separation, consistency management, read/write separation with sharding, and judicious use of NoSQL.

Application Architecture

Application architecture translates business capabilities and processes into concrete IT systems, defining required applications, their functions, and interactions.

Two Levels

Application Function Diagram : high‑level view of systems and functional modules, focusing on “what”.

Single‑Application Technical Diagram : internal layers (data access, business logic, API, front‑end) and their interactions, focusing on “how”.

Design principles: prioritize stability, decouple core from non‑core, enforce loose coupling (asynchronous calls, timeouts), and ensure fault tolerance through independent deployment and multi‑zone redundancy.

Technical Architecture

Technical architecture addresses non‑functional requirements such as availability, performance, scalability, security, and elasticity, selecting technology components and deployment patterns.

Three Perspectives

Physical Deployment : hardware, network, and cluster layout.

Logical IT Architecture : number of database, messaging, cache, and application servers.

Technology Stack : frameworks, languages, and middleware for storage, logic, and presentation layers.

Core principles: statelessness for horizontal scaling, reusable services (e.g., FineDataLink’s data asset and API capabilities), and governance mechanisms (degradation, rate limiting, feature toggles, monitoring).

Conclusion

All four layers are interdependent. Effective digital transformation requires starting with a clear business architecture, then sequentially realizing data, application, and technical architectures, ensuring the final design aligns with enterprise needs and can evolve.