OmniScribe

Context

Built with Google AI Studio — leveraging Google's speech and language models.

Problem

Raw audio recordings — meetings, interviews, lectures — are difficult to search, summarize, or act on. Existing tools perform poorly on Hebrew audio.

Solution

OmniScribe ingests audio files, transcribes them using state-of-the-art speech recognition optimized for both English and Hebrew, then applies NLP to extract key topics, action items, and summaries.

Screenshots

Tech Stack