HARVEST

Inspiration

Helping everyone to make money out of farming from what they have currently

What it does

Provide a recommendation for the type of crop and type of fertilizer based on where you live, what time of the year it is, and your goals (short term or long term profit)

How we built it

Typescript, React, CSS, FastAPI, Pandas, Python

Challenges we ran into

1. Scope of data was too large for the smaller scope of the hackathon
2. Scope of idea hence was also too large for the smaller scope of the hackathon
3. Libraries that were previously relied on for ML development were depreciated and we had to find new libraries that were compatible while performing the same function
4. Front end developers did not know how to use react or typescript and used tutorials to learn them and develop the frontend

Accomplishments that we're proud of

1. Successfully connected a front end to back end through APIs
2. Learned how to use React and Typescript
3. Developed a program that genuinely could be deployed for use in different communities

What we learned

1. We should research ML libraries beforehand and make sure they are not depreciated to save time/effort
2. We should come in with a detailed idea of the languages we plan to use and make sure everyone on the team is comfortable with it
3. Spend more time creating a P0 (MVP - Minimum Viable Product)
4. Better communication from the backend team so both backend and frontend are clear about the arguments that have to be inputted and their units so last minute changes do not have to be made

What's next for HARVEST

1. Formatting the output for the backend into a better usable format for the frontend can flush out more of the UI
2. Expand the scope to a global market predictor, as the project was meant as a POC (Proof of Concept) for this large scale predictor.

Built With

• ai
• api
• csv
• fastapi
• pandas
• python
• react
• typescript