Omi AI Mentor
Live transcription and concise summaries for Omi users, delivered as notifications.
Created on 2nd April 2025
•
Omi AI Mentor
Live transcription and concise summaries for Omi users, delivered as notifications.
The problem Omi AI Mentor solves
Omi AI Mentor enhances the Omi app by providing real-time transcription and concise, summarized responses. It captures live conversations within the Omi app, processes them on a backend server using an advanced language model, and summarizes key points with Transformers. These summaries are delivered as notifications to the Omi app, ensuring users stay informed with clear, actionable insights on the go. Omi AI Mentor addresses the challenge of processing and understanding conversations in real time for Omi wearable users. It simplifies communication by transcribing live audio, summarizing discussions, and delivering key insights as notifications. This helps users stay focused and informed without needing to manually review lengthy transcriptions, making interactions more efficient and actionable.
Challenges I ran into
A key challenge was achieving low-latency transcription processing to deliver real-time summaries. I faced a bug where network delays caused transcription lag, impacting the user experience. I resolved this by optimizing the backend with asynchronous API calls and implementing efficient polling, ensuring timely and accurate delivery of summarized responses to the Omi app.
Technologies used