This project is an English speaking practice assistant that leverages Google Gemini AI to evaluate pronunciation and provide real-time feedback. Users speak into a microphone and the system transcribes their speech, then uses Gemini to assess pronunciation accuracy, grammar, and provide correction suggestions. It also supports scenario-based conversation practice where the AI generates contextual dialogue situations. Optional ElevenLabs integration enables voice interaction. The system runs locally via a Python CLI application and requires a Gemini API key.