Transcription has been a valued service since the dawn of the written word. From the scribes of ancient Rome to secretary’s of the 1950s, civilization requires documentation of what happens around us. In the internet age, the realities and possibilities for transcribing content have changed.
TranscribeMe applies speech-recognition algorithms and a crowd-sourced platform to quickly provide accurate transcripts. First, the software makes audio and video content searchable and shareable by converting speech to text. Then, the text is divided up into 15-20 second pieces and doled out to a community of transcribers to process. By combining the power of technology and a human touch, the company seeks to accomplish both speed and accuracy.
Since August, the number of clients has tripled and companies like Ipsos, GigaOM, VMware, and Evernote have made use of the service. It can be used by individuals and businesses alike to manage their content.
“More people are telecommuting and self-publishing than ever before,” said VP of Sales Bethanie Maples Krogstad in a Q&A. “Corporations need call transcripts so they can analyze customer service issues. Writers need transcripts so they an spend more time thinking. Doctors need transcripts so they can spend time seeing patentees, not writing notes. Researchers everywhere need fast an highly accurate text of the complex audio produced from panels and interviews.”
Founders Alexei Dunayev and Greg Feerer got the idea for TranscribeMe while studying for their advanced degrees and wishing for a more efficient way to transcribe thesis interviews. While there are other transcription services out there such as GRM Transcription, Mechanical Turk, SpeechPad, TranscribeMe distinguishes itself with its hybrid approach to the issue.
“All other competitors force users to give up one aspect of value: turnaround time (speed), price, confidentiality, or quality (accuracy),” Krogstad said. “In terms of transcripts for professionals and corporations, either they have to have work transited in house, or pay a high-price for confidentiality.”
TranscribeMe launched in 2011 and has 2,500 transcribers on the platform. The team recruits workers from around the world and has raised an angel round to grow this workforce, as well as its internal team of 22. The funding of $900,000 came from Tech Coast Angels, Sierra Angels, Keiretsu, Maverick Angels, TEC Ventures, and ICE Angels.
VB's research team is studying web-personalization... Chime in here, and we’ll share the results.