From 1a6d6132dd04a40f95e41d349c5ca0cfd11b1cfb Mon Sep 17 00:00:00 2001 From: Bryant Date: Fri, 29 Nov 2024 01:57:08 +1100 Subject: [PATCH] improve docs --- fern/providers/transcriber/assembly-ai.mdx | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/fern/providers/transcriber/assembly-ai.mdx b/fern/providers/transcriber/assembly-ai.mdx index 531a8de..a0ab28e 100644 --- a/fern/providers/transcriber/assembly-ai.mdx +++ b/fern/providers/transcriber/assembly-ai.mdx @@ -11,7 +11,7 @@ AssemblyAI is a leading provider of AI-driven speech recognition and understandi **The Evolution of AI Transcription:** -Speech recognition has evolved from basic systems to sophisticated AI models capable of understanding diverse languages and accents. AssemblyAI has been at the forefront of this evolution, developing models like Universal-1, trained on over 12.5 million hours of multilingual audio data, achieving best-in-class accuracy across multiple languages. +Speech recognition has evolved from basic systems to sophisticated AI models capable of understanding diverse languages and accents. AssemblyAI has been at the forefront of this evolution, developing models like Universal-2, trained on over 12.5 million hours of audio data, achieving best-in-class transcription accuracy across several industry-critical languages. **Overview of AssemblyAI's Offerings:** @@ -19,15 +19,15 @@ AssemblyAI offers a comprehensive suite of AI-driven tools designed to meet dive **Speech-To-Text** -- Their core offering converts spoken language into written text with high accuracy, supporting real-time transcription and batch processing. +- Their core offering converts spoken language into written text with up to 95% accuracy and 30% less hallucinations than other leaders in the space. **Audio Intelligence** -- Beyond transcription, AssemblyAI's models can analyze audio to detect sentiment, identify topics, and perform speaker diarization, providing deeper insights into conversations. +- Beyond transcription, AssemblyAI's fully featured speech understanding models can analyze audio to detect sentiment, identify topics, and perform speaker diarization, transforming words into meaningful ideas, insights, and opportunities. **Real-time Transcription** -- AssemblyAI's real-time transcription feature enables instant conversion of speech to text, beneficial for live captioning, customer support, and interactive voice response systems, enhancing user experience and operational efficiency. +- AssemblyAI's real-time transcription feature enables sub-second latency conversion of speech to text, beneficial for live captioning, customer support, and interactive voice response systems, enhancing user experience and operational efficiency. **Use Cases for AssemblyAI**