Tech Stacks¶

A tech stack is the set of languages, frameworks, databases, and infrastructure used to build a system. This page is a compact reference — for architectural patterns and layering, see Architecture.

The Layered View¶

A modern web system is usually composed of these layers. Each layer has many viable choices; the list below is representative, not exhaustive.

Layer	Purpose	Common choices
Edge	CDN, DDoS protection, DNS	Cloudflare, CloudFront, Akamai, Fastly
Presentation	User interface	React, Vue, Angular, Svelte, Flutter, Swift, Kotlin
API / Integration	Service contracts	REST, GraphQL, gRPC, WebSocket
Business Logic	Application frameworks	Spring Boot, Django, FastAPI, Express, Rails, .NET
Messaging	Async communication	Kafka, RabbitMQ, SQS, Celery
Data Access	ORMs, search, cache	SQLAlchemy, Hibernate, Sequelize, Elasticsearch
Storage	Persistent state	PostgreSQL, MySQL, MongoDB, Redis, S3
Infrastructure	Runtime and orchestration	AWS, GCP, Azure, Docker, Kubernetes

Languages at a Glance¶

Language	Sweet spot	Notable frameworks
Python	Data, ML, scripting, web APIs	Django, FastAPI, Flask
JavaScript / TypeScript	Frontend, full-stack, Node backends	React, Vue, Next.js, Express, NestJS
Java	Enterprise, Android, high-scale services	Spring Boot, Quarkus
C#	Windows, enterprise, game dev (Unity)	.NET, ASP.NET Core
Go	Cloud infrastructure, microservices, CLI	Gin, Echo
Rust	Systems, performance-critical services	Actix, Axum
Ruby	Rapid web development, scripting	Rails, Sinatra
PHP	Traditional web, CMS	Laravel, Symfony

Databases at a Glance¶

See Architecture § Databases for the full breakdown. Quick picks:

Default choice: PostgreSQL. Relational, ACID, JSON support, mature.
Document store: MongoDB for flexible schemas; DynamoDB for serverless AWS.
Cache / queue: Redis — in-memory, versatile.
Analytics: BigQuery, Snowflake, ClickHouse.
Wide-column at scale: Cassandra, Bigtable.

Classic Stack Combinations¶

Stack	Components	Typical use
LAMP	Linux + Apache + MySQL + PHP	Traditional web apps, CMS
MEAN	MongoDB + Express + Angular + Node	SPAs, real-time apps
MERN	MongoDB + Express + React + Node	Modern SPAs
JAMstack	JavaScript + APIs + Markup	Static sites, blogs, JAMstack apps
Django + PostgreSQL + Redis	Python web + relational + cache	Content and data-driven apps
Spring Boot + PostgreSQL + Kubernetes	Java + relational + orchestration	Enterprise microservices

How to Pick a Stack¶

Order of considerations, roughly in this priority:

Team expertise. A familiar stack ships faster than a theoretically better one.
Problem fit. ML-heavy? Python. High-throughput trading? Rust, C++, or Java. Real-time collab? Node or Elixir.
Ecosystem maturity. Libraries, hiring pool, Stack Overflow answers, long-term viability.
Operational cost. Managed services (RDS, Cloud Run) beat self-hosting when you can afford them.
Future-proofing, carefully. Don't adopt yesterday's fad; don't chase tomorrow's either.

Common mistake: picking the stack before understanding the problem. Write the one-page problem statement first.