This comparison examines iii-engine, a modern service composition framework, against the traditional Express.js + Postgres stack that has dominated backend application development for over a decade. The contrast highlights fundamental architectural differences in how contemporary serverless and edge-native systems approach infrastructure composition, data persistence, and operational requirements.¹⁾

Express.js + Postgres represents the conventional approach to building web services. Express.js is a minimal HTTP framework for Node.js that handles routing, middleware, and request-response cycles, while PostgreSQL provides relational data storage. This combination typically requires additional supporting infrastructure: SSE (Server-Sent Events) or Socket.io for real-time communication, pm2 for process management, external vector databases like pgvector for embeddings, Prometheus for monitoring, and containerization via Docker for deployment consistency.

iii-engine takes a fundamentally different approach as a service composition framework that integrates multiple operational concerns into unified primitives. Rather than assembling separate tools, iii-engine provides built-in abstractions for functions, key-value storage, streams, and OpenTelemetry (OTEL) instrumentation as first-class primitives. This design eliminates the need for external databases, separate monitoring systems, or container orchestration.

The Express.js + Postgres stack requires careful orchestration of multiple independent systems. Each component handles a specific concern: Express manages HTTP routing and middleware chains, Postgres handles persistent state and complex queries, Socket.io or SSE manage bidirectional communication, pm2 handles process lifecycle and clustering, Prometheus scrapes metrics from instrumented endpoints, and Docker packages everything into reproducible containers. Integration between these systems requires explicit configuration, network setup, and operational overhead.

iii-engine consolidates these concerns through primitive abstractions. Functions provide computation without requiring separate HTTP framework configuration. Key-value storage replaces the need for external databases—eliminating schema design, migration management, and database administration. Streams enable event-driven communication patterns without Socket.io or SSE complexity. OpenTelemetry integration is built-in, eliminating separate monitoring infrastructure. This unified approach reduces configuration surface area and deployment complexity significantly.

Express.js + Postgres deployments typically follow traditional container-based patterns. Services run in Docker containers orchestrated by Kubernetes or similar platforms. Database operations require connection pooling, backup strategies, replication configuration, and capacity planning. Process management via pm2 handles restarts and clustering. Monitoring requires separate Prometheus installations with configuration, alerting rules, and dashboards. This approach scales well for large teams and complex requirements but incurs operational overhead.

iii-engine eliminates many operational concerns by providing infrastructure as part of the framework itself. No external database means no database administration, schema migrations, or connection management. No Docker requirement simplifies deployment to serverless and edge environments. Built-in OTEL support means observability is instrumented without separate tool installation. This approach is particularly advantageous for smaller teams, rapid prototyping, and edge computing scenarios where traditional infrastructure becomes unwieldy.

PostgreSQL offers powerful relational data modeling with ACID guarantees, complex query capabilities, and transaction support. Developers design schemas explicitly and use SQL to query across relationships. pgvector extension enables semantic search over embeddings. This flexibility comes with schema design costs and query optimization requirements.

iii-engine's key-value primitive provides simpler state management without schema constraints. This approach suits applications prioritizing rapid iteration and simple data structures over complex relational queries. For applications requiring semantic search or embeddings, iii-engine's KV store integrates with OTEL instrumentation, providing alternative patterns for similarity-based retrieval.

Express.js + Postgres remains optimal for applications requiring complex relational queries, large-scale team coordination, mature operational practices, or existing PostgreSQL investments. It provides proven stability and extensive ecosystem tooling.

iii-engine excels in scenarios emphasizing minimal operational overhead: edge computing, serverless deployments, AI agent memory systems requiring distributed state, rapid prototyping, and applications where built-in observability and unified primitives reduce configuration burden. Its approach aligns with modern cloud-native and AI-centric application patterns.

• Service Composition Framework (iii-engine)
• iii-engine
• iii.dev