Livesearch is an (unfinished) Next.js application specifically designed to provide contextual content generation based on audio clips. By leveraging OpenAI's GPT-4 and Whisper, it generates context-based notes and other content to aid in the conversation.
When you're interested in context-based notes, you can simply record your audio (auto-mode coming soon), which will be automatically transcribed by Whisper. Then, GPT-4 uses the transcription as well as some engineered prompts to generate notes. Other forms of content are coming soon.