Audio Search Engine

Objectives

• Upload and index a library of audio files with transcriptions
• Build a full-text search index over all transcribed content
• Display search results with context snippets and timestamps
• Click a result to play the audio starting at the matched timestamp
• Support filters by date, duration, and custom tags

Hints

• Use Whisper to batch-transcribe all audio files with timestamps
• Store transcripts in a searchable index using Fuse.js or a database
• Map each search result back to a specific time offset in the audio

Resources

• platform.openai.com/docs/guides/speech-to-text
• www.fusejs.io

Comments (0)