Core Technologies
- ASR (Automatic Speech Recognition)
Automatic speech recognition is the technology to extract the transcript of human speech in the given language. It requires a huge amount of human-labeled training data in different scenarios including domains, channels, and dialects. With such training data, AI Rudder optimizes ASR models tailored specifically for financial industry where the accuracy in its scenarios such as customer acquisition, underwriting, reminder, and collection is significantly better than alternatives in the market. The accumulated advantage through utilizing the feedback from its customers allow AI Rudder to maintain market-leading position as ASR vendor and help customers success in improving business procedure using AI.- NLU (Natural Language Understanding)
Natural language understanding is focused on helping computers understand, interpret, and manipulate human language data. AI Rudder is experienced in handling data in different languages, even some data in mixed languages. Such handling includes data cleaning, transforming, clustering, and labeling by AI Rudder's scientists, linguists, and engineers. Combined with a set of simple to sophisticated models, these data can be trained into state-of-the-art intelligent modules that work for the products and services AI Rudder has to offer.- TTS (Text-To-Speech)
Contrary to ASR, Text-to-speech transforms texts into speech. To inject intonation into machine- generated speech, AI Rudder innovates a hybrid synthesizer that combine human recorded voice and machine computed one. Elements such as numbers, dates, and people's names can be generated in real time to broaden the possibilities of covering different business scenarios. All of these are crucial for AI Rudder's robot to interact with end users in a meaningful way with human touch brought about by AI.