EchoQuest

Challenge

Extracting key insights from hours of recorded interviews is a manual and time-consuming task.

Solution

I designed 'EchoQuest,' an end-to-end pipeline that processes raw audio into a searchable knowledge base. It enhances audio, separates speakers (diarization), transcribes with OpenAI's Whisper, and indexes the content.

Result

A powerful analysis tool where a user can ask natural language questions about the interview ('What did Sarah say about the budget?') and get instant, accurate answers.