Development of AI assistant for local user in Python programming language

Abstract

The article considers a system for developing a multifunctional AI assistant capable of performing a wide range of tasks. An AI assistant model developed based on neural networks is presented, which provides high accuracy in executing commands and is capable of working in real time. The proposed model has achieved high accuracy in voice recognition, working effectively even when used without an Internet connection. The model integrates the Vosk and Speech Recognition speech recognition systems, which allow for accurate and efficient processing of voice commands. Voice commands are recognized in real time, regardless of whether they are simple commands for individual actions or scenarios that include obtaining information. Text input processing for executing commands is also implemented. Particular attention in the work was paid to the development of an effective mathematical model that minimizes delays in executing requests and ensures stable operation.