KeepNotesV2

A production-oriented Notes API built with FastAPI, Pydantic, and native async PyMongo.

This project is primarily a backend engineering personal project focused on:

• FastAPI architecture
• Pydantic validation
• MongoDB and query design
• Authentication and JWT security
• REST API design
• Async Python
• Testing
• Security best practices
• Production-oriented architecture

Features

• User authentication and authorization
• JWT-based authentication
• Password hashing with Argon2
• CRUD operations for notes
• Note visibility:
  • Private
  • List
  • Friends
  • Public
• Note status:
  • Draft
  • Published
  • Archived
• Tag-based note searching
• Public note exploration/feed
• Cursor-based pagination
• MongoDB aggregation pipelines
• Admin analytics
• User note statistics
• Filtering and sorting
• Async database operations
• Pydantic request/response validation

Pagination

The project implements cursor/keyset pagination for feed-style endpoints.

Instead of relying on increasingly expensive offsets such as:

page=500&limit=10

the API uses a cursor based on a deterministic ordering:

created_at + _id

The cursor is encoded into a token and returned to the client as next_cursor.

The client treats the cursor as an opaque value and sends it back when requesting the next batch.

Conceptually:

GET /explore?limit=10
        ↓
   10 notes
        +
   next_cursor
        ↓
GET /explore?limit=10&cursor=...
        ↓
   next 10 notes

This project also explores how pagination strategy connects with MongoDB indexing and query design.

Feed / Explore

The /explore endpoint demonstrates a simple feed architecture:

published + public notes
        ↓
   deterministic sort
        ↓
  cursor pagination
        ↓
      client

The project also uses this as a foundation for understanding more advanced feed systems such as candidate generation, ranking, personalization, and recommendation systems.

MongoDB

MongoDB is used as the primary database through native async PyMongo.

The project covers:

• MongoDB queries
• Filtering
• Sorting
• Projection
• Aggregation pipelines
• $group
• $lookup
• $project
• $arrayElemAt
• Pagination
• Query/index design

Tech Stack

• Python
• FastAPI
• Pydantic
• MongoDB
• PyMongo Async
• JWT
• Argon2

Project Goal

KeepNotesV2 was built primarily as a backend engineering learning project.

The goal was not simply to build a notes application, but to use the application as a vehicle for learning how real backend systems are designed:

• API architecture
• Database modeling
• Query design
• Authentication
• Security
• Pagination
• Indexing
• Aggregation
• Scalability considerations
• Production-oriented design decisions

Status

✅ Project completed

KeepNotesV2 is considered complete as a learning project.

Future experiments involving real-time communication, social graphs, personalized feeds, and WebSockets will be explored in separate projects.