Backend

Overview

Eggy Backend runs a NestJS application as a single AWS Lambda behind API Gateway, with MongoDB as the primary data store via @nestjs/mongoose. Deployments are handled by Serverless Framework on AWS, executed from CircleCI, which is triggered by a GitHub Action.

• Single Lambda simplifies deployment and infra while supporting REST APIs.
• MongoDB + Mongoose provides flexible schemas and rapid iteration.
• CI/CD: GitHub → CircleCI → Serverless deploy to AWS.

Request lifecycle (Lambda)

1. Client calls REST endpoint on API Gateway.
2. API Gateway invokes the Lambda (NestJS app).
3. On a cold start, the Nest app boots, modules initialize, and a MongoDB connection is established.
4. The request is routed to the appropriate controller/route handler.
5. The response is returned via API Gateway to the client.
6. On warm starts, the same Lambda process is reused, reducing latency and reusing the DB connection.

Cold starts (basics)

• Runtime: Node.js 20 on arm64 is a good default (lower cost, good perf).
• Bundle size and dependency count affect cold starts; keep the package lean.
• Memory size affects CPU during init; more memory can reduce cold start time.
• Keep DB connections and app instances reusable between invocations when possible.

Where to next

• See Architecture for details of the Lambda/Nest integration and packaging.
• See Data Modeling for @nestjs/mongoose schema guidance.
• See Local Development for running locally with MongoDB (Docker) or Atlas.
• See CI/CD → Serverless Deploy for the deployment pipeline and environments.