Framework planning for overall program

Inspiration: Our inspiration is our strong feeling about giving back to the community and helping those who do not have the means or ability to live a healthy life. As Boy Scouts, we have strong feelings about food drives and giving back to the community and we all understand the struggles of asking people for canned goods and the challenges of getting donations from homes. This program allows homeless shelters and banks which tap into an unused market that has even more potential of being able to support our communities.

How it works A big issue of current society is that the people who have a lot are unable or do not have the means to share with those who need it. For example, dairy is a food that has a hard cut-off time for its exploration date. Many people do not buy milk within a week of its expiration date, so unfortunately, it will go unsold and uneaten. However, within this timeframe, it can still be eaten which is where we come in. We aim to connect food banks and shelters that can use these foods within a usable timeframe which in other cases would be wasted.

How we built it We all first came up with an idea based on topics that we were all close to. One problem that resonated with us was the fact that so many homeless people are unable to access food. We built this app using HTML/CSS/Javascript as our frontend and a backend using Python, the backend included programs to clean the data, query an LLM based on a CSV file comprising of data of all the stores, and run a FastAPI server which the website queries. We used a wireframe to come up with these capabilities initially and then implemented it using VS Code and Github.

Challenges we ran into

Problems with ensuring structured output from the LLMs we used in the AI-enabled grocery Store Search
Storing data coming from multiple sources in structured formats which can both be read/written to by both the backend and frontend
Integrating backend servers with frontend technology to ensure that no matter what user input is, the cleaned output is being delivered (ex. AI-enabled Search, FastAPI Calls, etc.)
Some issues we faced on the front end were dealing with the extensive lines of code and complicated features however

Accomplishments that we're proud of We are all proud of the finished result because of the elegant website, complex code, and direct response to problems in the community. Our overall idea is something we all feel strongly about and our program actively attempts to respond to the problem in a unique way which is not currently being done.

** What we learned** Abhay exercised his programming prowess and worked on people skills and teamwork abilities, he improved his ability to work with RAGs and implement models like ChatGPT with dynamic websites. Aryan learned how to apply his app-building skills to a website format and got to practice his HTML skills. Arjun learned website building skills from Arjun and he got to practice frontend dev. Aditya learned the ins and outs of git and git hub and learned website-building practices from both Aryan and Abhay. This executes the diverse skill sets among our teammates. It accentuates our ability to teach others and split tasks based on which skill sets best fit the job.

What next In our current form, we have a proof of concept that accurately shows that our program can analyze and connect companies so that food can be unwanted in communities. However, with more time, we would be able to further this program by allowing a larger range of data to be accepted, and a more accurate and knowledgeable understanding of how long foods can be used past expiration date while keeping in mind FDA guidelines and consumer health.