The great duality of the internet is that, despite its immense modern sophistication and the fact that billions of people use it, many parts of it are still rather similar to its early days.
Articles and newsletters are probably the best example, with published works on the internet in the 90s still working in pretty much the same way as they do now – text on the page. While certain things have changed with nicer looking websites, different business models and media embeds, the core premise is the same.
It’s not often that such an engrained part of media is shaken up by something that improves it so dramatically. The big buzzword of internet culture is always disruption, but what if we could take what already exists and just make it better?
SpeechKit is doing just that with its sophisticated text-to-speech solution that takes those tried and tested written articles or newsletters and provides a voice to them by way of synthesised speech.
The first benefit of this is, of course, accessibility. With traditional online media struggling to attract and keep readers, offering an audio option opens the content up to the widest array of people. Not just those with difficulty seeing or reading, but an audio format also means those “sit down and read” articles become “take with me on the go” articles. Exercise, driving and many other activities now become avenues for people to keep reading, or should we say listening.
But anyone can throw together some code that will churn out audio in a monotone overseas accent that is a pain to listen to. What makes SpeechKit special is the lengths it goes to perfect speech-to-text and make its solution the best on the market.
It uses Speech Synthesis Markup Language (SSML) and Natural Language Programming (NLP) modules to make speech that is familiar and natural to listen to.
You can take a dive into this interesting work with SpeechKit’s augmentation of Amazon Polly to make better audio articles. As an example of this it uses South African president Cyril Ramaphosa as a local name that may be difficult for overseas code to properly pronounce.
Representing local voices
And South African context isn’t just used as a one off example mentioned above. You may have already come across SpeechKit’s work out in the wilds of local news.
News24.com has been using SpeechKit to bring audio to its articles for some time now, offering a perfect local case study of an existing publication adding this functionality to its offering seamlessly and without a major teardown and rebuild of the existing site.
Even more exciting is that this isn’t limited to South Africa, the United States or the UK. Because of the customisability of this solution it can be tailored to anywhere you want to sound anyway you want. In beta right now is even a feature called Clone which will allow users to create a synthetic version of their own voices, so that they themselves can relay content to readers.
If your business, publishing house, newsletter or blog could make use of SpeechKit, signup for a free 7-day trial. For a demo of enterprise features, including API integration, high-quality custom voices and 24/7 support, get in touch with one of their team.