Portfolio: CardTurk

During the development of Enable Labs' MORI product, we identified that our clients were sitting on hundreds of thousands of index cards containing old member records. Some of these organizations are hundreds of years old, and have their history documented in such a way that a fire or flood could erase it forever.

After considering and rejecting the possiblity of OCR for the cards (inconsistent formats, handwriting, etc.,) we developed CardTurk as a means for our clients to recruit volunteers from within their organization to transcribe information from the cards into database records with the human intelligence required to truly understand the cards.

Since CardTurk's launch in early 2013, more than 7000 cards have been transcribed by a pool of volunteers for just the first client using it.

Although CardTurk was conceived for the use of existing clients, it has been built to facilitate human transcription of any set of images where OCR is determined to be impractical or human intelligence is needed to understand the images.

CardTurk Screen Shot

Technologies Used

  • Languages:
    • Ruby (Ruby on Rails, HAML, Capybara, RSpec)
    • Javascript (Twitter Bootstrap, jQuery)
  • Development Tools:
    • Sublime
    • Mac
  • Deployment:
    • Linux
  • Database
    • MySQL
    • File System (for card images)


Newsletter Sign-Up

Be the first to find out about the goings on at the Lab!