Transcription services typically convert spoken words into a written or digital document. While this job was once reserved for secretaries or personal assistants (PAs) skilled in shorthand and touch typing, technology has automated the task to make it more accessible to a wider audience.
For example, as a copywriter I conduct interviews with clients to get the details about their business and the message they want to portray. With the recorded interview, I can use transcribing software to determine the individual words on the audio track and output a text document to add to my reference files. In reality, however, many transcribing systems need considerable training to deliver real accuracy, and even then, their capabilities are limited to the voice they have been trained to recognize.
Tropo provides an API for automating communications by connecting code to a phone network for both voice and messaging. This service is able to transcribe any recording, including multi-party conference calls. This tutorial by Adam Kalsey on the Tropo blog guides followers through creating advanced transcription with analytics using Tropo with VoiceBase’s audio indexing and transcription API.
Using Slim Framework, Tropo, and VoiceBase, the small application described in this tutorial sends a Tropo recording to VoiceBase for transcription using an API call. The application then asks for a basic machine transcription, waits for the results, and then performs the designated action, such as emailing the text file or storing it in a database.
The entire application consists of fewer than 100 lines of code, all of which is provided, along with comments on what each line does. The application will run on any PHP web server, and the complete app is available on GitHub.