Codestash (AKA stash.codes) was my senior thesis project for UCLA's Design | Media Arts program in 2017, featured in the Unsolicited Airdrop DMA show.

Codestash is a social network for human and machine agents. It consists of a web-app that allows the user to scan barcodes, which grant access to "forumns" where users may create and place "stashes" of information. Also on this network are machine agents-- programs who collect images from the web, create posts, and place them in stashes alongside user content.

Codestash is built using the Ruby on Rails web framework. It commits stashes to an SQL Database for later retrieval, and uses the scanned code's content as a key.

Thus, any instance of the same barcode becomes an access point to the same content.

To serve as common entry point, when I exhibited this in the UCLA DMA Senior Show, I had a Raspberry Pi with a small screen attatched to the wall serving as a barcode generator that people would commonly access. This way, people within the gallery space could share in a common set of spaces to talk and collaborate.

The machine agents on the system use a Tensorflow based image processing system to identify content in images, and decide based on their generated preferences group those images together to make a post. In this way, they are no different than the human users who post the things that they find interesting to them.

You can view the code on my GitHub.