Inspiration
I was inspired by the repetitive challenge of summarizing long,** multilingual meetings**, where context switching and precision were critical. I wanted a tool that lets me stay focused during discussions while automating the administrative burden afterward.
Focus of the project
I was required to attend meetings and report to the stakeholders and development team regarding the minutes of meeting, status, progress, and action items for multiple projects every week. Initially, it took around 1 hour 30 minutes to prepare and send updates via email threads with relevant CCs.
I had to re-listen to the entire meeting (which often lasted an hour) and manually summarize key items — this process was not only time-consuming but also mentally draining.
How we built it
I envisioned a tool that acts like a meeting companion — intelligent enough to:
Highlight key moments in real time
Record personal notes on command
Automatically generate personalized summaries and follow-up emails
For this, we used langgraph agents, webhooks , python for backend tasks. For frontend, we used bolt, along with typescript and vite build
Challenges we ran into
The discussions were multi-lingual — often switching between English and local languages.
Key points were mentioned sporadically, so I had to manually identify timestamps and extract important details within 10-minute intervals.
It was difficult to stay engaged and take live notes while listening carefully for critical details.
What the App Does
Highlight Key Points, send meeting notes as mail or chat to selected members via app in any format as mentioned.
Using a shortcut (⌘ + I), the user can mark an important moment.
The system captures the timestamp and transcript segment around that highlight.
Take Voice Notes Instantly
Press (⌘ + G) during the meeting → it records your short voice note, transcribes it, and stores it in a Notes section.
Automated Meeting Minutes
Challenges Faced (Previously)
The discussions were multi-lingual — often switching between English and local languages.
Key points were mentioned sporadically, so I had to manually identify timestamps and extract important details within 10-minute intervals.
It was difficult to stay engaged and take live notes while listening carefully for critical details.
Future Scope:
Integrations: Gmail API, Teams API, Slack API Security: OAuth 2.0 + KMS Encryption Analytics: Read.ai API or Custom ML Pipeline
Built With
- amazon-ec2
- amazon-web-services
- api-gateway
- claude-sonnet
- langgraph
- microsoft-teams-developement
- openai
- python
- s3
- websockets
- whisper
Log in or sign up for Devpost to join the conversation.