AzBuki.ML

Technologies such as voice assistants, chatbots, smart home devices, "artificial intelligence" in various forms, etc. are being increasingly integrated in human’s daily life. The connection between us and technology is deepening, and computers are no longer controlled only by a mouse and keyboard, but by sound, gestures, commands, and language. Leading companies and organizations in the field of technology already have well-developed machine learning platforms for processing unstructured natural language, which are used in products such as Siri, Google Assistant, Translate and IoT devices, which can now talk to us and respond to our commands. However, many of these human interactions with technology through language are supported mostly in English and a small group of several other languages. Although there is support in Bulgarian portals for translation or sound recognition, there aren’t mainstream practical platforms offering models for classification, grammatical analysis, content analysis, etc. To improve on that we propose our solution - “АзБуки.ML”, an innovative machine learning platform for natural language processing focused on Slavic speech. Its main goal is to efficiently implement technologies such as recurrent neural networks and advanced NLP algorithms in the IT business by providing easy-to-use APIs for third-parties while simultaneously having powerful first-party tools.