Mastering the Architect Role: Responsibilities, Mindsets, and Frameworks

Architect Responsibilities

Architects lead teams, define and confirm requirements, decompose systems into overall architectures, make technology selections, write specifications, and drive implementation.

Typical Roles

Enterprise architect – focuses on overall IT architecture.

Software product architect – concentrates on product development.

Application architect – customizes solutions, covering overall, application, data, and deployment architecture.

Technical architect – handles infrastructure, cloud platforms, product recommendations, deployment, networking, and data‑center design.

Solution Architect Work

Understand client pain points, define high‑level and non‑functional requirements, map client assets (e.g., StarRing, Alibaba Cloud), iterate proposals, deliver overall architecture, and make architectural decisions.

Responsibilities from Different Views

From the client: build confidence, solve mid‑level problems, lead technical adoption.

From the project: coordinate with management, PMs, business, development, testing, operations, and configuration teams.

From Alibaba internal: support sales solutions, market promotion, capture customer needs, and consolidate solutions.

Architectural Thinking

Combines top‑down and bottom‑up approaches, emphasizes problem definition, first‑principles analysis, metric modeling, and philosophical thinking (e.g., evolution, entropy, fractals) to drive innovation.

Top‑Down Architecture

1. Define the problem, especially the client’s core issue. 2. Add a time dimension to separate means from goals. 3. Use layered and dimensional thinking to uncover true needs. 4. Prioritize solving the client’s mission before internal goals. 5. Apply multiple analysis methods to translate business problems into required capabilities. 6. Refine processes and capability models. 7. Define and decompose metrics, then model mathematically. 8. Convert abstract capabilities into technical challenges for bottom‑up design.

Bottom‑Up Application Architecture

Derive system sequence diagrams from business processes, aggregate modules into larger components, and build the system architecture through four sub‑paths: business concept architecture, system model, system process, and non‑functional support (performance, stability, cost).

Domain‑Driven Design (DDD)

Analyze user scenarios to identify use cases, create robust diagrams to discover entities, partition domains, and evaluate the partitioning for optimal design.

Data‑Driven Design

In the era of IoT, big data, and AI, shift from pure domain analysis to data‑statistical analysis and BI, using big‑data insights to shape business, application, data, and technical architectures.

Common Architecture Frameworks

Key frameworks include:

TOGAF – The Open Group Architecture Framework, with ADM (Architecture Development Method), content framework, reference models, and capability framework.

Zachman – Organizes architecture using six questions (what, how, where, who, when, why) across six layers.

ITSA – Early HP enterprise architecture framework focusing on essential components.

DODAF – U.S. Department of Defense framework with eight viewpoints (AV, CV, OV, SvcV, SV, DIV, StdV, PV).

Architectural Principles

Design principles guide component decomposition and dependency management:

OCP – Open/Closed Principle.

SRP – Single Responsibility Principle.

LSP – Liskov Substitution Principle.

ISP – Interface Segregation Principle.

DIP – Dependency Inversion Principle.

REP – Reuse‑Publish Equivalence.

CCP – Common Closure Principle.

CRP – Common Reuse Principle.

Practical guidelines include N+1 design, rollback strategy, feature toggles, monitoring (traceId, business ID), multi‑active data centers, mature technology adoption, resource isolation, horizontal scaling, buying non‑core components, commercial hardware, fast iteration, and stateless design.

Typical Architecture Evolution

Progression from single‑server monolith to distributed systems:

Separate application, data, and file services.

Introduce caching (local and distributed).

Cluster application servers behind load balancers.

Database read/write separation with master‑slave replication.

Reverse proxy and CDN acceleration.

Distributed file systems and databases.

NoSQL and search engines for complex queries.

Business‑level service splitting (vertical and horizontal).

Distributed services (cells) with dedicated partitions.

Architecture Management

Adopt a win‑win model and result‑oriented management to align stakeholder expectations, ensure project delivery, and measure architectural outcomes.

References: Alibaba internal articles, industry blogs, and technical books.