Software Engineering

Application Architecture by Use Case

Architecture that adapts to what you are building — not what is fashionable. Context-aware decisions, not organisational habit.

The Problem

Why "One Architecture Fits All" Fails

Organisations standardise on a single architecture pattern, a single stack, or a single deployment model and force every product into it. This leads to over-engineered simple systems, under-controlled critical systems, unnecessary cost, poor developer experience, and fragile operations.

Clavon avoids this by anchoring architecture decisions to use case characteristics, not organisational habit.

Selection Criteria

Architecture Selection Matrix

Before selecting an architecture, we classify the application across these axes:

User Profile

External users, internal users, or both?

Interaction Style

Read-heavy, write-heavy, real-time, batch?

Data Sensitivity

Regulated, financial, personal, low-risk?

Change Frequency

Daily iteration or quarterly releases?

Scale Driver

Users, transactions, data volume, integrations?

Integration Density

Few integrations or many upstream/downstream systems?

Failure Impact

Inconvenient, costly, or catastrophic?

Application Types

Architecture by Use Case

Six distinct application contexts, each demanding a different architectural approach.

01

Web Applications

Characteristics

  • -Browser-based access
  • -Mix of read/write operations
  • -High UX expectations
  • -Variable traffic patterns

Recommended Architecture

Modular monolith (default) with clear domain modules, BFF where UX complexity justifies it, REST APIs for client communication.

Common Failure Modes

  • -Premature microservices
  • -UI logic leaking into backend
  • -Overuse of synchronous integrations
  • -No performance budget

Clavon guidance: Optimize for change speed first. Introduce async patterns only when justified.

02

Mobile Application Backends

Characteristics

  • -Mobile clients are unreliable (networks, latency)
  • -API contracts must be stable
  • -Backward compatibility is critical
  • -Auth and authorization are central

Recommended Architecture

API-first backend with clear versioning strategy, stateless services, modular monolith or microservices depending on team maturity.

Common Failure Modes

  • -Breaking API changes
  • -Chatty APIs that drain battery
  • -Overloaded endpoints
  • -Poor error semantics

Clavon guidance: Mobile clients move slower than servers. Backends must be conservative and backward-compatible.

03

SaaS Platforms

Characteristics

  • -Multiple customers share infrastructure
  • -Strong isolation requirements
  • -Subscription, billing, entitlements
  • -Long-lived data
  • -High availability expectations

Recommended Architecture

Modular monolith or carefully designed microservices. Key decisions: tenant isolation model, data partitioning strategy, configuration vs customization boundaries.

Common Failure Modes

  • -Tenant data leakage
  • -Hard-coded tenant assumptions
  • -Over-customisation per customer
  • -Billing logic scattered across services

Clavon guidance: Treat multi-tenancy as a core domain. Design isolation explicitly. Plan migration paths between isolation levels.

04

Enterprise Internal Systems

Characteristics

  • -Complex workflows
  • -Many user roles
  • -Heavy ERP/CRM integration
  • -Compliance and audit requirements
  • -Predictable but critical usage

Recommended Architecture

Modular monolith with strong boundaries, explicit workflow orchestration, integration layer abstracting ERP/CRM dependencies.

Common Failure Modes

  • -Business logic embedded in integrations
  • -Tight coupling to ERP schemas
  • -Poor test coverage for workflows
  • -No ownership clarity

Clavon guidance: Internal systems should reduce complexity, not mirror it.

05

API-First & Integration-Centric

Characteristics

  • -No UI or thin UI
  • -Heavy upstream/downstream dependencies
  • -Contract stability is critical
  • -Failures cascade quickly

Recommended Architecture

API gateway + core services, strict contract governance, async events for state changes, synchronous APIs for commands.

Common Failure Modes

  • -No versioning strategy
  • -Shared databases
  • -Hidden coupling via payloads
  • -Poor observability

Clavon guidance: APIs are products. They require ownership, documentation, and lifecycle management.

06

Data-Intensive & Analytics-Driven

Characteristics

  • -Large volumes of data
  • -Batch and streaming workloads
  • -Complex transformations
  • -Downstream AI or reporting consumers

Recommended Architecture

Separate transactional and analytical paths, event or CDC-based data ingestion, clear lineage and ownership, scalable storage and compute separation.

Common Failure Modes

  • -Analytics impacting transactional performance
  • -Unclear data ownership
  • -Manual data fixes
  • -No monitoring of data quality

Clavon guidance: Never mix operational workloads and analytics without strong boundaries.

Cross-Cutting Concerns

Decisions That Apply Across All Use Cases

Authentication & Authorization

  • Centralised identity
  • Context-aware authorization
  • Token scopes and claims
  • Explicit service-to-service trust

Error Handling & Resilience

  • Timeouts and retries defined explicitly
  • Circuit breakers for integrations
  • Graceful degradation strategies
  • User-visible vs internal error separation

Observability

  • Logs, metrics, and traces per use case
  • Business-level events captured
  • SLIs aligned to user impact
Evolution

Architecture Evolution Triggers

Traffic growthScale-out specific modules
Team growthExtract ownership-aligned services
Compliance scope increasesStrengthen controls and logging
Integration sprawlIntroduce integration layer
Performance bottlenecksSeparate hot paths
Anti-Patterns

Common Anti-Patterns Eliminated

Reusing the same architecture blindly across different use cases

Designing for hypothetical future scale before the problem is real

Overusing async patterns where synchronous is simpler

Ignoring operational reality during design

Treating architecture as done rather than continuously evolved

Artefacts

Deliverables

Use-case-specific reference architectures

Decision rationale per application type

Integration and scaling strategies

Quality and operability implications

Architecture evolution roadmap with triggers

Related Services

This Practice Directly Supports

Start a Conversation

Ready to Build Context-Aware Architecture?

Clavon anchors every architecture decision to use case characteristics — not organisational habit or fashionable patterns.