AWS Case Study: Magnify.net
Steve Rosenbaum is the founder and CEO of Magnify.net, a video publishing platform that makes it easy for customers to integrate user-generated videos, videos they produce, or videos they discover into their Websites. He is also the author of the book Curation Nation.
At the outset of his book project, Rosenbaum realized he would be interviewing, writing, and editing on an extremely tight schedule. He needed the power of a distributed workforce to help write the book: “As an author, I've got to be able to manage more data, more files, and more types of media. And that requires me to have two things: cloud support and a team of human assistants I can engage on a moment’s notice.”
Why Amazon Web Services
Rosenbaum looked to AWS (Amazon Web Services) for that support. He explains, “AWS solutions became the backbone of my publishing process, giving me a new way to work quickly and with a standard of quality that my publisher and my readers expect.”
To leverage the power of AWS, Rosenbaum created two workflows: one for local interviews and one for interviews to be conducted remotely. He explains how the process worked: “For interviews I conducted face to face, it was all about video. For example, when I interviewed Newscorp's Chief Digital Officer, the entire hour was recorded in HD video. Then, using Magnify.net, I uploaded the interview as a series of segments. Here, Amazon Simple Storage Service (Amazon S3) provided cloud storage and delivery to our CDN."
Once those files were up, Rosenbaum set up an Amazon Mechanical Turk Human Intelligence Task (HIT) for the 15 interview questions, reaching out to Workers in the Amazon Mechanical Turk marketplace with a high approval rating to do audio transcriptions. The work was done quickly and professionally. Next, transcripts were brought in to a Microsoft Word document and scanned for the best thoughts and ideas.
“When I couldn't be face to face with a subject, I did the interviews on Skype, and recorded the audio. Those files were cut up, put on Magnify.net, and sent to Amazon Mechanical Turk for transcription," says Rosenbaum.
Rosenbaum spent less than $350 on transcripts for the book; using other methods, this service would have cost thousands of dollars and taken much longer.
Knowing that Amazon Mechanical Turk Workers participate for both money and the opportunity to be part of something interesting, Rosenbaum made an effort to share with them information about the interview subjects, what they were talking about, and how this information would become part of his book project.
Rosenbaum most appreciates that he was able to speed up the parts of the production process that had previously been a bottleneck, leaving time for the careful and deliberative part of the book-making process. He notes, “Being able to interview, transcribe, and write in near real time was a liberating experience. The result was a project that has its roots in a carefully constructed business book, but also has video, audio, and transcripts as digital assets that historically wouldn't have existed.” He adds, “The future of the book is multi-format, with the shape and size morphing to fit the shape and size of the device and the needs of the reader.”
Learn more about how AWS can your company with work that requires human intelligence by visiting our Mechanical Turk details page: http://aws.amazon.com/mturk/.