I had almost no time for this data engineering challenge, because the early invite messages were sent out on the 25th of April and the final cut-off date was the 29th. I therefore chose to focus on two feature-sets:
- Adding photos to a database quickly, in a scalable way, and in aggregate
- Viewing photos added to the database
I implemented these requirements with a React frontend and an AWS backend, using an S3 bucket as a data repository. This has several advantages, but most significantly it results in high availability and concurrency, and is well-suited for hosting large media files like images.
All environment variables are as follows (only available on Notion version of README).
The .env file has to be in the project’s root directory, and it must have the exact prefixes shown above to work (create-react-app limits the syntax of environment variables to avoid exposure).
Install all project dependencies with npm install.
To run the application, navigate to the root directory of the project and type npm start into the terminal. This will create a development build which exists at [localhost:3000](http://localhost:3000) by default.
There are many, many things that I wish I had time to add before the deadline. Here are the top ones:
- A robust caching system and CDN, probably built with Cloudflare
- A DynamoDB database attached to the S3 bucket which generated queryable metadata programmatically when new information is added into it
- Topic labelling for incoming images using SageMaker and probably the ImageNet classification algorithm
- The integration of pre-signed URIs for data security without the need for a backend.
- Automatic refresh of the application upon