Inspiration
We were inspired by the difficulty of swapping between different LLMs and receiving varying responses between them.
What it does
SuperAI is a website that takes in a prompt from a user and then sends it to four different LLMs. It then uses a machine learning algorithm to decide on the best response using a variety of attributes, and also create an aggregate response.
How we built it
We Built SuperAI using Svelte, SvelteKit, Tailwind, Typescript, Groq, Cloudflare, Python, and FastAPI.
Challenges we ran into
We ran into challenges in making a user-friendly interface accessible through different devices. We also ran into challenges in creating a fast and reliable back-end to support our AI model.
Accomplishments that we're proud of
We are proud of implementing the APIs for four different LLMs in a single application, as well as the back-end AI model that aggregates through the LLMs and selects the best-fitting response.
What we learned
We learned concepts of API's as well as the deployment process.
What's next for SuperAI
More features
Built With
- ai
- cloudflare
- css
- css3
- fastapi
- groq
- html
- html5
- huggingface
- javascript
- python
- svelte
- sveltekit
- tailwind
- transformers
- ts
- wolfram-technologies

Log in or sign up for Devpost to join the conversation.