Are you curious to get to know the team behind Klang.ai? Today Mattias Fält, doctor in mathematics and co-founder of Klang, answers Andreas's short four questions about Klang and the technology behind the service.
Are you curious to get to know the team behind Klang.ai? Today Mattias Fält, doctor in mathematics and co-founder of Klang, answers Andreas's short four questions about Klang and the technology behind the service.
That's a good question!
The foundation of all these AI models is built on mathematics. It's a collection of a lot of numbers in a large network that you need to train to get better at what you're supposed to do. Whether it's generating speech or generating text or generating an image, everything is built on a lot of numbers.
In mathematical optimization, it's about finding the right numbers to make something as good as possible.
This fits well with what you're trying to do with these AI models. You're trying to make the model as good as possible, which means finding the right numbers.
Klang's AI model is specifically trained to be good at Swedish, rather than being generally good at all languages. This has been achieved by training the model on a lot of audio and text in Swedish.
So AI models are trained by showing them examples of audio and corresponding transcribed text. By repeating this process millions of times, the model becomes better at generating text from audio.
The biggest challenge is to always deliver high-quality transcriptions, even when audio quality is poor, people talk over each other, or have difficult dialects.
Want to listen to the entire interview? Take a look at the interview in the service 👈
Did you know that this entire text was built with our functionality for finding questions and answers from interviews? Try it yourself on Klang.ai ⭐