Kieli provides fully labelled datasets of Arabic language for machine learning, including text and audio data for natural language processing. The data is aimed at enhancing artificial intelligence and machine learning systems with support for speech and text analysis.