CompliantWasteAI

• 

  UIprototype

• 

  UIprototype

Inspiration

After working in the Medical Waste field as an IT specialist, I've witnessed the inefficiencies and the high frequency of federal and state fines, even practice closure. Th cost of standard medical waste consulting is unjustifiable considering that a large language model and data curation. Not only does this contribute to inefficiency, the lack of compliance of these regulations hurts our environment already tarnished by man-made emissions.

What it does

We utilize sentence embedding and vectorize approved FDA and EPA data to determine the legal classifications of pharmaceutical waste, industrial waste, and medical waste. The user can use natural language to describe their waste in detail to receive the most crucial information to schedule a pickup from a certified vendor.

As we are beginner Computer Science Community College students, we will include the work of pressed pill recognition within the platform, accessible UI, and In-app ordering of pickups with available schedule times.

How we built it

We used Sentence Bert models, pineconeDB and OpenAi's API to build the core of this app. We prototyped the computer vision aspect with Pytorch and YOLOv8. For the in-progress UI, we are using Flask

Accomplishments that we're proud of

We're simply proud of building the core of this app. back-to-back communication with non-technical experts and the lack of experience was conquered within hours. We know that given extra hours, we are capable of building a product ready to ship, helping the environment and American businesses.

What we learned

We all have the collective knowledge of introductory computer science classes given our computer science program at the community college we attend lacks comprehensive curriculum for the prerequisites of this project. Yet we were determined and spent countless hours learning how to integrate LLMs and vectors into our code.

What's next for CompliantWasteAI

We hope to expand on the code given a week, continue our communication with chemists and waste experts, as well as submit the project for review to the FDA for nation-wide approval.

Built With

• python