Skip to main content

Specialized Domains: Architecture Considerations

Overview​

Different problem domains demand different architectural approaches. A real-time trading system has fundamentally different constraints than an e-commerce platform. An IoT device with kilobytes of memory faces challenges that a data center with terabytes doesn't. Machine learning systems introduce novel concerns around model serving, drift, and experiment tracking. Compliance-heavy systems require audit trails and segregation of duties.

This section explores 10 specialized domains where architecture decisions matter most:

Real-Time Systems require deterministic latency, hard deadlines, and prioritized scheduling. Think embedded controllers, autonomous vehicles, and industrial automation.

Streaming and Event Processing handle unbounded data flows with exactly-once semantics, watermarks, and backpressure. Kafka, Flink, and Spark Streaming enable this.

IoT and Edge Computing balance constrained devices (limited CPU, memory, power) with cloud connectivity. OTA updates, gateway patterns, and device management are critical.

Machine Learning Systems introduce feature engineering, model serving, training/serving skew, concept drift, and the data science lifecycle alongside software engineering.

High-Compliance Systems (HIPAA, PCI-DSS, SOX) require immutable audit logs, segregation of duties, access controls, and regulatory reporting—not just security.

Gaming Systems optimize for minimal latency, state synchronization across players, prediction and extrapolation, and massive concurrent connections.

Embedded and Firmware manage bare-metal constraints, interrupt-driven execution, memory-constrained algorithms, and hardware abstraction layers.

Fintech and Payments ensure idempotency, reconciliation, double-entry bookkeeping, fraud detection, and settlement correctness—where failures cost money.

E-Commerce orchestrates catalog management, cart/checkout flows, inventory allocation, dynamic pricing, and promotion engines at scale.

Social and Communication builds feeds (fan-out on write vs read), real-time messaging, content moderation, abuse prevention, and notification systems.

Each domain has unique constraints, tradeoffs, and patterns. This section provides the architectural vocabulary, decision frameworks, and practical examples to guide you.

Learning Map​

How specialized domains relate and their key concerns

Prerequisites​

Before diving into specialized domains, you should be comfortable with:

  • Architectural Fundamentals: Layers, decomposition, interfaces, APIs
  • Distributed Systems Basics: CAP theorem, eventual consistency, message-driven communication
  • Quality Attributes: Performance, reliability, security, scalability
  • Design Patterns: Observer, Command, State, Strategy
  • Data Architecture: Databases, caches, queues, topics
  • Observability: Logging, metrics, distributed tracing

Articles in This Section​