OmiSphere

Inspiration

All of us have access to very limited information, so we end up making semi-random decisions. We are here to change it

What it does

We analyze the data from the users conversations to create a collective consciousness. Then, we create personalised knowledge graphs to help each user understand themselves better, their peers and the whole world

How we built it

We connect the omi device through elevenlabs speech-to-text service to our backend which uses an LLM to extract topics, sub topics, insights and then plots the knowledge graphs - first with your interests, 2nd with people who have similar passions, third with general world insights

Challenges we ran into

setting up the pipeline (from OMI to the front end graph building)

Accomplishments that we're proud of

Making it work Collaborating in the amazing team

What we learned

you should build with amazing people, listen to each other, have fun, incorporate the feedback

What's next for OmiSphere

world-wide distribution of users, multi-omi networking

Built With

• elevenlabs
• nextjs
• shadcn
• supabase
• trpc
• typescript
• vercel