uniscan

Demo Video

View the demo video here.

Installation

Clone the repository

git clone https://github.com/hwennnn/uniscan.git
cd uniscan

Copy the environment variables

cp .env.example .env

cd backend
cp .env.example .env

cd ../

Run Docker compose

We will be running the app in development mode for now.

docker-compose -f docker-compose.yml -f docker-compose.dev.yml --env-file .env up --build

Setup Prisma Schema (in development)

After spinning up the containers, run the following command to initialise the database schema:

# SSH into the container
docker exec -it uniscan-nestjs bash

# Inside the container shell
pnpm db:push # to push the schema to the database
# or
pnpm db:sync

After pushing the DB schema, you are ready to go 🎉🎉.

View the frontend at http://localhost:5173 and the backend at http://localhost:8000/api/v1.

API Documentation

Please refer to the API documentation for more information.

Architecture Considerations

Backend Architecture

For the backend, the following technologies were chosen:

• NestJS
• Prisma
• PostgreSQL
• Redis
• BullMQ
• ethers
• Typescript

For the backend, I am using NestJS, Prisma, PostgreSQL, Redis, and BullMQ to effectively handle high transaction throughput which is required in this project. I am also more familiar with NestJS, Prisma, and PostgreSQL so that helps a lot to complete this in a short amount of time. I choose PostgreSQL as it's a powerful relational database that offers robust data integrity and supports complex queries. Redis plays a crucial role in caching frequently accessed queries such as the latest ethereum price and integrates seamlessly with BullMQ to distribute workload across the queue. By utilizing a message queue, I have decoupled the processing of historical transaction batches from the main server operations. Jobs are executed through the queue, and the client continuously polls to check the status of these jobs, ensuring a more efficient workflow.

Frontend Architecture

For the frontend, the chosen technologies include:

• Vite
• Tailwind CSS
• Typescript
• React Query

For the frontend, I am using Vite alongside React Query. React Query provides query optimization and offers numerous built-in features for managing complex queries and pagination.

Run Testing

Backend

pnpm test

docker exec -it uniscan-nestjs bash

# Inside the container shell
pnpm test:e2e

WIP

To setup configs to run the production environment

docker-compose -f docker-compose.yml -f docker-compose.prod.yml --env-file .env up --build