EEGSpeech. A brain-computer interface (BCI) for decoding speech phonemes from EEG signals using a hybrid CNN-LSTM model, with interactive Streamlit visualizations and Docker support.