Clarq
About the Project
We built Clarq to make it easier for students, professionals, and creators to capture and retain important information from lectures and meetings. Taking notes can be distracting and time-consuming, and it’s easy to miss key points.
Clarq listens in real time, recognizes different speakers, transcribes and organizes the content into clean, searchable notes, and even suggests follow-ups, reminders, or helpful resources based on the discussion. This context-aware AI helps users not just record information but also take meaningful next steps.
How We Built It
- Speech-to-Text: Uses pre-trained models to transcribe audio in real time
- Speaker Identification: Distinguishes who is speaking during discussions
- Note Structuring: Organizes transcriptions into clear sections and key points
- Context-Aware Suggestions: After each session, the AI recommends follow-ups and action items
Our team split tasks to focus on backend AI, frontend design, and testing, collaborating closely to improve the overall experience.
What We Learned
- Integrating multiple AI tools for real-time audio processing was challenging but rewarding
- Team collaboration strengthened testing, debugging, and refining the product
- Making the AI context-aware showed how notes can guide users’ next steps, not just record information
Challenges
- Multiple speakers: Separating voices when people talk over each other
- Real-time performance: Ensuring notes are generated fast enough for live sessions
- Summarization & guidance: Condensing speech into readable notes while giving useful follow-up suggestions
Built With
- amazon-web-services
- deepgram
- electron
- javascript
- node.js
- node.js-**apis:**-deepgram
- openai
- openai-gpt
- react
- typescript
- typescript-**frameworks:**-react
- vite
Log in or sign up for Devpost to join the conversation.