Definition
Instead of one monolithic codebase, microservices separate concerns — auth service, billing service, recommendation service — each with its own repo, database, and deploy cycle.
Trade-offs: team autonomy and scaling vs. operational complexity and distributed debugging.
In simple terms
A monolith is one large store where every department shares a back room. Microservices are specialty shops that coordinate by phone — flexible, but more coordination overhead.
Where you see it
- Netflix and Spotify popularized microservices at scale.
- ML inference may run as a separate model-serving microservice.
How it works
1.Bound contexts
Split by business capability, not arbitrary layers.
2.APIs between services
REST, gRPC, or message queues.
3.Deploy independently
CI/CD per service with versioned contracts.
Why it matters
- Understanding microservices helps you navigate modern cloud-native backends.
Often confused
Microservices are always better than monoliths.
Many successful products start monolithic; split when team or scale pain justifies the cost.