SpeechTrainer
Aplikacja stworzona, aby poprawić sposób oraz jakość nauki nowych języków. Zawiera specjalne moduły sztucznej inteligencji oraz automatyzuje tradycyjne metody oceny kursantów.
What did we have to face?
The most important issues that we had to solve
Speech recognition
One of the main challenges was the necessity to develop an AI algorithm that effectively distinguishes different accents and dialects within a single language. It is important that the application can be used by people from various regions who may pronounce words in slightly different ways.
Customization of the AI algorithm
The application had to be designed in such a way that it could evaluate the pronunciation of both beginner and advanced users. This meant the necessity of creating an assessment system that can recognize minor mistakes of beginners, as well as those of more advanced users.
Real-time analysis
The challenge was to ensure that speech analysis was conducted in real time, without delays, which is crucial for the interactivity of the application. This required the optimization of both AI algorithms and server infrastructure to ensure smooth application performance even under heavy load.
How did we solve this?
The technologies used that helped us achieve our product with the highest quality
Phone
