New AI Platform Decodes Emotions in Speech with Unprecedented Precision

A new platform has expanded its capabilities to analyse emotions in spoken language. By combining speech-to-text processing with advanced speech analysis, it now extracts emotional insights from audio and video data. The system was developed through partnerships with leading audio technology firms. In June, the platform introduced a speech-to-text feature created with AssemblyAI. This tool provides automated transcription and tags emotional markers in spoken content. AssemblyAI’s software also includes speaker identification, sentiment analysis, and keyword extraction.

The platform has also launched a speech analysis module in collaboration with audEERING. This module uses audEERING’s devAIce platform, which is built on openSMILE 3.0—a widely used open-source toolkit for speech and emotion recognition. Together, these features allow rapid and precise emotional analysis of interviews, lectures, and customer interactions.

The combined capabilities of the platform are designed to support fields like healthcare, education, and marketing. They also extend to mental health, user experience research, and communication studies. The platform’s integrated tools now enable deeper exploration of human behaviour through voice and speech analysis. Businesses, researchers, and healthcare providers can use these features to gain insights from audio and video data. The technology is already being applied across multiple industries.