Inspiration Our inspiration came from the challenges students face when preparing for meetings, study groups, or even casual conversations. Many students struggle to organize their thoughts and often hesitate to ask questions because they're unsure how to articulate them. We wanted to create a tool that helps students gain clarity and confidence by turning their scattered thoughts into coherent summaries.
What it does ClearSpeak allows students to upload audio or video recordings of their thoughts, meetings, or discussions. The app transcribes the audio, asks clarifying questions, and then generates a clear, structured summary or personalized report. This helps students prepare for meetings, organize their thoughts, and feel more confident in their communication.
How we built it We built ClearSpeak using Whisper AI for speech-to-text transcription and LangChain for prompt engineering and iterative clarifying questions. The front end is developed in React and TypeScript, while the back end uses Supabase for real-time data updates and file storage. We focused on creating a smooth user experience, allowing users to interact with the app easily and effectively.
Challenges we ran into One of the biggest challenges was ensuring the AI could generate accurate clarifying questions that truly enhance the transcription. Another challenge was integrating the speech-to-text functionality smoothly and efficiently, given the varying quality of audio files. Lastly, managing file sizes and ensuring fast processing without losing accuracy was a significant hurdle.
Accomplishments that we're proud of We’re proud of creating an intuitive app that takes audio input and turns it into something useful and structured. Our team successfully implemented the clarifying questions loop, which ensures that the output is personalized and contextually accurate. We’re also proud of the smooth integration between our front-end and back-end systems, ensuring a seamless experience for users.
What we learned We learned the importance of iterative development and refining our prompts to get the best output from AI models. We also gained valuable experience in balancing technical performance (speed and accuracy) with user experience, especially when working with audio and real-time transcription.
What's next for ClearSpeak Next, we plan to introduce more customization options, allowing users to personalize their summaries even further. We’re also looking to add visual representations like diagrams for enhanced organization and understanding. Additionally, integrating with tools like Zapier and supporting real-time collaboration are future goals to expand ClearSpeak's functionality.
Log in or sign up for Devpost to join the conversation.