SoapBox Labs said its new speech recognition API will be free for developers to use for the next three months.
SoapBox Labs has launched a new API to help companies developing educational and entertainment tools and apps for kids to quickly and securely integrate voice technology into their offerings.
The news comes shortly after the Irish speech recognition tech start-up announced a €5.8m funding round.
SoapBox Labs said it is releasing the API to support literacy and learning at home, while parents around the world seek out new remote learning tools for children.
Designed with developers of digital games, literacy and English language learning applications in mind, the plug-and-play API is being offered free of charge for an initial three-month period.
Dr Patricia Scanlon, CEO of SoapBox Labs, said: “With this API, we’re offering companies a secure and straightforward way to enrich children’s digital experiences using voice technology.”
The three-month free API offering will incorporate a range of support services to help developers who are integrating voice technology into children’s education and entertainment tools for the first time. These services include demos, developer documentation and access to in-house experts.
The technology has already been adopted by Lingumi, an English language platform for kids. Lingumi CEO Toby Mather said that integrating SoapBox’s tech has “driven consistent engagement and helped us to improve the quality and consistency of our learning engine”.
SoapBox Labs said that most clients can integrate the API themselves within 24 to 72 hours. The start-up also offers APIs that perform more standard automated speech recognition and fluency measurements, which involve more engagement with the client.
Niamh Bushnell, chief communications officer at SoapBox Labs, explained: “Developers get access to our API, start hitting the server from a command line to test the system themselves and then they integrate it.
“It’s the simplicity that makes our API a plug-and-play experience. It’s designed to be quick and easy with no customisation, near real-time responses, and it’s globally scalable.”
How it works
Bushnell added that literacy and language apps using SoapBox Labs’ technology typically work by prompting a child to read or recall a word, sound or sentence.
“The app records the audio, sending it via a simple command to our restful API. Our API responds in near-real time – less than 500 milliseconds – with a confidence score showing how well the child pronounced the sound, word or phrase down to the phonetic level,” she said.
“The developer can then decide how they would like to feed that response back to the child, the parent and/or the teacher.”
Bushnell explained that Lingumi does this by displaying a thermometer of measurement for the child to see how accurate their English pronunciation is, which encourages the kids to keep working on getting a higher score to raise the thermometer measurement.
“We’ve also seen traffic lights, word highlighting and sounds used in apps to communicate accuracy and progress back to the child,” she added.
“If the developer chooses, audio can be played back to the child along with scores, or the scoring can be invisible. As I say, this is at the discretion of the developer and it depends on the type and purpose of the app they’re building.”
Learning from home
SoapBox Labs said that many parents are looking for alternatives to “passive” learning and literacy apps, to find suitable materials that will engage and immerse kids and allow them to learn more independently.
However, it added that there are also privacy concerns when it comes to the platforms that children are using to play and learn online. Many existing tools do not provide transparent details on the data they gather from children or how they use that data for profiling and advertising.
SoapBox Labs said that it advocates for children’s data privacy and has taken a proprietary “privacy-by-design” approach to building its product.