Overview
Kudos is a peer recognition platform: employees nominate a teammate by category and skill, a reviewer approves or rejects, and approved recognitions award stars that show up on profiles, the directory and the leaderboard.
I designed and built the complete backend from scratch in two days using AI-assisted development with Cursor and Claude — while owning the architecture, database design and API behaviour end to end.
How it works
- An employee signs in with an organisation email (validated against the allowed domain) and receives a JWT.
- They nominate a teammate, choosing a category and a skill.
- Reviewers work through a review queue and approve or reject each nomination.
- Approved recognitions award stars that appear on profiles, the people directory and the leaderboard.
My contribution
- Designed and built the backend from scratch: custom users, recognition workflow and admin APIs.
- Delivered organisation-email login, JWT with blacklist logout, review queues and dashboard counts.
- Built superuser management of users, skills and categories, documented with OpenAPI.
- Met a two-day deadline using Cursor and Claude without compromising on structure or code quality.