This is a submission for the AssemblyAI Challenge : Sophisticated Speech-to-Text & No More Monkey Business.

What I Built

VoiceLoom is an AI-powered voice notes application that transforms stream-of-consciousness recordings into structured, actionable content. Powered by AssemblyAI's LeMUR model, it tackles the universal challenge of converting unstructured thoughts into organised, valuable information.

Key Features

• Intelligent Speech-to-Text: Crystal-clear transcription using AssemblyAI's Transcription API
• Smart Summarization: Concise summaries via the summarisation parameter in Transcription API
• Structured Output: Transformation of fuzzy thoughts into organised content using LeMUR
• Dynamic Title Generation: Automatic creation of contextual titles using LeMUR
• Action Item Detection: Smart extraction of action points from voice recordings

Demo

Experience VoiceLoom: Launch App

Note: Bring your own AssemblyAI API key, configurable in settings after login.

Visual Journey 📸

Welcome to VoiceLoom

Your gateway to organized thoughts

Voice Capture Interface

Simple, intuitive recording experience

Structured Notes View

Your thoughts beautifully organized

Source code

Journey

After exploring AssemblyAI's comprehensive API documentation, VoiceLoom emerged as the perfect showcase for the platform's capabilities. The project addresses two key challenge prompts:

1. Sophisticated Speech-to-Text

   • Advanced transcription with precise punctuation
   • Intelligent summary generation

2. No More Monkey Business

   • LeMUR-powered title generation
   • Structured note formatting
   • Smart action item extraction

Universal-2 Implementation

• Generates professional transcriptions with proper punctuation
• Creates informative summaries from voice content

LeMUR Integration

• Dynamic title generation based on content analysis
• Conversion of unstructured thoughts into formatted notes
• Intelligent action item identification

Future Roadmap 🛣️

1. Interactive Note Analysis

   • Chat interface for note exploration
   • LeMUR-powered Q&A using transcript IDs

2. Enhanced Insights

   • Automated note highlights
   • Intelligent pattern recognition
   • Personalised content analysis