Dropbox is adding a new feature to its platform that aims to help users identify text within billions of image and PDF files.
Using machine intelligence, Dropbox said it's built AutoOCR, an automatic image text recognition capability that performs optical character detection on photos of documents, including receipts, field reports and more. The system is able to recognize the files so that the docs will surface when users search within Dropbox.
"Looking for a contract that a teammate scanned years ago? Just search for the vendor," Dropbox wrote in a blog post. "Trying to track down blueprints an architect put together for your remodel? Type in their name. Can't remember where you saved that flight itinerary screenshot? Enter in the destination airport."
Dropbox said this was one of the most computationally intensive projects that the company had ever attempted, and credits machine learning for allowing the feature to operate at scale. More broadly, AutoOCR is part of the Dropbox intelligence initiative (DBXi), Dropbox's effort to incorporate AI into all of its products and services.
"By scaling our internal machine intelligence platform, DBXi is multiplying the efforts of our dedicated intelligence product team so all our engineers can modify and validate models for intelligent features, improved search, and other business optimizations," Dropbox wrote.
Dropbox said the AutoOCR update will launch today and roll out to users over the coming weeks.
PREVIOUS AND RELATED COVERAGE:
Dropbox expects to have a quarter of its Magic Pocket infrastructure on SMR drive capacity by 2019.
Dropbox becomes the latest company to embrace AMD EPYC and see a benefit to single-socket platform adoption.
The new updates include team selective sync, member space limits, directory restrictions, and the ability to disable downloads of shared links.