EdenAI

Inspiration

The inspiration for this project comes from a time one of our members attempted to garden during the summer and the plants died

What it does

EdenAI takes parameters from a user and finds plants that would likely be part of their ideal garden. There is also a chatbot that helps the user with questions and gives advice on how to grow this theoretical garden into reality.

How we built it

The front-end was built with React and Next.js. The back-end was made using typescript, react, and firebase. The AI used was a custom PyTorch model and a Gemini API.

Challenges we ran into

The main challenge we ran into was bug fixing. Whenever one bug was fixed another was created. Many hours and 90% of our brainpower was spent trying to deal with this. We also faced difficulties in connected the different parts of our project together, e.g. the ai model created using PyTorch with the rest of the website.

Accomplishments that we're proud of

Something that we are proud of is being able to get some sort of functionality with many new frameworks despite having no experience. This was a good learning experience for us and we hope to use the knowledge we gained today in the future.

What we learned

Throughout the hackathon we have learnt many new frameworks as we had essentially no real practical experience with them, e.g PyTorch. However, another important thing we learnt was the importance of being coordinated with your team. At the start progress was slow due to a lack of coordination which only really picked up near the end.

What's next for EdenAI

The next step for EdenAi is improving our functionalities. One, our custom AI could use more training to handle a wider variety of situations. Two, our Gemini API needs better implementation to provide better results of different plants people can use in their garden. Three, improving scalability so that more people can access a new part of their life, both for themselves and the future of our planet.