Inspiration

We were using aws s3 to store files while building a data platform. Had to download csv, excel's inorder to view the content during debugging. Troubles increased when we switched to parquet files (cannot be opened by standard softwares).

What it does

Accesses AWS S3 and provides a simple interface to browse tthe files and allows to view contents of tabular files like csv, excel, parquet in a data grid. JSON is pretty printed and images are also supported.

How we built it

We used a client server architecture with our web app built in flutter web hosted on firebase hosting and our flask api made in python hosted on heroku.

Challenges we ran into

Initially, we wanted just a client side application, but due to technical constraints when dealing with parquet files as well as lack of aws sdk for flutter, we decided to go for a flutter - python architecture.

Heroku dyno's (free plan) sleep after 15 minutes of inactivity but to avoid any inconvenience to out users, we hit a warmup api endpoint whenever our web app loads.

What's next for S3ViewerSR

1) Adding support to render html content. 2) Rendering ipynb file format (Jupyter notebook)

Share this project:

Updates