CulturAI

CulturAI is an explorative platform for users to deepen their understanding of local communities through a fun, interactive experience

Comment

Inspiration: We are living in an increasingly interconnected world and this has resulted in greater polarisation and fragmentation. We wanted to bridge this divide in a fun, engaging manner. As we are also avid travellers who enjoy experiencing new cultures, we came up with CulturAI for users to explore communities through a fun, immersive experience.

We thought of incorporating a scavenger hunt style theme by introducing riddles for mystery locations to allow users to have a fun explorative experience. The places in the list are places of activities so users can experience the lives of the local community. Some activities include local festivals, trying out local food, music festivals, exploring an ancient site, etc.

What it does: CulturAI is an explorative platform for users to deepen their understanding of local communities through a fun, interactive experience. Users have to solve LLM- generated riddles to unlock mystery locations in the community they want to explore.

How we built it: Next js for frontend, flask for backend, ngrok for tunnelling, Mistral AI 7B, Controlnet for image to image synthesis, ResNet with embeddings for image classification.

Challenges we ran into: One challenge we faced was coming up with data to train our models for places of unique cultural experience. As these places are often places locals would be aware of, it was challenging to get them online.

Accomplishments that we're proud of: We managed to build an mvp incorporating LLM in such a short time.

What we learned: We learnt how to incorporate AI elements to our project to make it fun and to understand the user journey in-depth. By having a comprehensive understanding of user experience, we can better design the UI/UX.

What's next for CulturAI: We could bring this technology to every rural community in the world and scale it, bridging different cultures in the world

Note that the images below are Figma mockups which can be developed upon

Built With

  • classification.
  • controlnet-for-image-to-image-synthesis
  • embeddings
  • flask-for-backend
  • for
  • image
  • mistral-ai-7b
  • next-js-for-frontend
  • ngrok-for-tunnelling
  • resnet
  • with
Share this project:

Updates