PeerPrep Background

PeerPrep is a Software Engineering project, the objective is to create a technical interview preparation platform with collaboration capability. Users are matched with a partner to solve the question provided by the platform. PeerPrep provides collaboration functionalites such as shared editor and video call.

Under the Hood

In this project, we adopted the Microservices architecture. Each service is built using Express.js. Our frontend portal is a React application using Next.js and Chakra UI. Depending on the service’s functionality, we used either PostgreSQL or MongoDB for the database. For matching and real-time functionalities, RabbitMQ and Web Sockets are used.

All of the services are consolidated under an Nginx API Gateway. They are deployed using Docker containers with Kubernetes as the orchestrator. We have a pipeline with Github Actions that deploys it to AWS Suite hosting — Elastic Container Registry, Elastic Kubernetes Service.

Project Insights

This assignment was a great opprtunity to learn and try out different technologies. It was fun working with Nginx, Docker and Kubernetes for the first time. I remember learning about auth_request directive in Nginx and refactored the project’s authentication flow to make it more elegant and streamlined.

Porting all the services from Docker Compose to Kubernetes was challenging but it felt really fulfilling when I managed to incorporate functionalities like Horizontal Pod Autoscaling and Self-healing.