July

AI came firmly into my life because I started creating a prototype of a speech recognition application for Android. So far, I have implemented both Speech-to-text and Text-to-speech.

Speech-to-text is powered by Vosk and Text-to-speech by Sherpa-Onnx. Both libraries work on-device without sending data to a server.

August

I'll integrate an LLM between STT and TTS to enable the application to provide voice consultations based on the provided dataset.

P.S. I publish weekly reports in the TG channel.