AssemblyAI, a leader in AI-based audio intelligence, has unveiled a series of new capabilities designed to help businesses extract and analyze insights from digital conversation data. The move comes as businesses increasingly turn to AI solutions to manage the vast amounts of information generated by virtual meetings, call centers, and chatbots.
Conversational Intelligence AI
Conversational Intelligence AI is quickly gaining traction as an essential tool for navigating the flood of digital conversational data. According to AssemblyAI, their platform offers a variety of features that maximize the value of audio data. Key features include:
- Sentiment analysis: This feature detects the emotion of each spoken sentence in scripted text, providing insight into the emotional tone of a conversation.
- Topic detection: This tool uses the IAB Content Taxonomy to identify different topics within a transcript, helping users categorize and understand the main topics discussed.
- Auto chapters: This feature summarizes audio data over time into multiple chapters, making it easier for users to navigate and find specific information.
- Key phrases: This tool identifies important words and phrases in transcripts to extract the most important concepts or highlights.
Rumor improvements
AssemblyAI has also introduced enhancements to its Large Language Model Usage Reporting (LeMUR) system. The latest update includes two new keys in the LeMUR response.input_tokens and output_tokens—This will allow users to track token usage more effectively. This add-on aims to help users manage their usage and keep it within desired thresholds.
AssemblyAI has also implemented spending alerts, allowing users to set up email notifications when their balance reaches a self-determined threshold. This feature further helps users monitor usage and manage costs.
New tutorials and resources
AssemblyAI continues to support the community with new tutorials and resources. Recent blog posts include guides on voice-to-text streaming, YouTube video transcription using Node.js, hotword detection with the top-level speaker segmentation library, and API exploration. These resources are intended to help developers and researchers leverage the capabilities of AssemblyAI in a variety of applications.
Additionally, AssemblyAI’s YouTube channel features popular tutorials such as Building an AI Voice Translator that Can Translate in 30+ Languages, Creating a Server-to-Server App to Transcribe Zoom Recordings, and Developing a Conversational AI with Real-Time Transcription Using LLAMA 3 There is this. Eleven Labs.
For more information about AssemblyAI’s new features and resources, visit our official blog.
Image source: Shutterstock