This was submitted for the Gen Z Insights track and the ElevenLabs Voice AI track at the ElevenLabs x Bolt.new London Founder House hackathon on 28/06-29/06.

Inspiration

Don't want to feel left out because you're busy? Don't want to catch up 500 unread messages from your mates WhatsApp group chat? Don't want to feel lost when they make plans?

Most of us skim read the messages and ignore the noise, but CC helps you stay in the loop by listening out for messages and consistently generating summaries on the fly so you never miss out on a message again.

What it does

CC summarises large quantities of incoming messages from DMs or group chats on any platform (WhatsApp, Instagram, etc) into short, actionable insights, filtering them down to the essence, and playing them back to you as if you were on a short group call in that chat or with that person - perfect for when you're on the go and wanting to understand what you missed out on!

How we built it

We initially used Bolt.new to generate a rapid prototype mockup for our frontend, and now we are using React Native to finetune and customise it for our needs so that we are able to easily deploy it on mobile. For our backend, we are using the FastAPI Python framework for our webserver, and in conjunction we use the OpenAI API to generate an "extract" for every summary and a full transcript for all the different ElevenLabs AI voices. Of course, for voice TTS generation, we use various custom ElevenLabs voices. Each user in the group chat has the ability to create their own voice, assuming that they are a user of the CC app too.

Challenges we ran into

Initially we thought it would be a difficult challenge to aggregate all messages received on various platforms e.g. WhatsApp/Instagram without using their API or pretending to be a "linked device" in the case of WhatsApp. However, we found a nice workaround by reading Android notifications as they come through, ensuring that we can remain platform agnostic and easily parse masses of messages at once.

Accomplishments that we're proud of

Developing a fully working full-stack app end-to-end, especially considering the spectrum of web dev experience that the team had, with some members having lots of experience & others having little to none.

What we learned

Combining all of our talents in the most effective way possible to build this app in a short amount of time. The power of version control in managing the codebase effectively, so each of us could work on our allotted segments of code in parallel and easily combine them.

What's next for CC

Incorporating MCP AI agent integrations with other apps such as Google Calendar or Notion Calendar, and linking CC with ElevenLabs' conversational voice assistant.

There would be huge scope to scale this app,

Built With

Share this project:

Updates