Countless news articles regarding lack of healthcare in many parts around the world inspired us to create this project.
Our project takes in user input about their symptoms and any additional concerns that they have. Using keywords, we find possible diagnoses and allow the user to choose what fits their symptoms best. Then, once the user selects their potential diagnosis, the website provides a step by step process allowing them to recover oftentimes without healthcare professionals.
How we built it
We used a mongoDB database to record user info, such as username, password, and most importantly recovery plan. We chose mongo since the DB structure was simple to follow and easy to understand.
Then, we used requests and json in python to scrape MayoHealth. We believe this is an essential step (passing in scraped requests to the LLM as context) to ensure no hallucinations are made in what could be potentially life-altering decsions.
We used Flask to display the results and take in user input. Flask is really fast to set up, and if needed, is able to handle users. For example, Netflix uses Flask. In this time crunch hackathon, we chose flask for speed.
Finally, we used HTML5, CSS, JS, and Tailwind to make the results nicely formatted and look vibrant (as well as the login pages). Aayush actually drew those himself! Parsing results with JS is continued in the next section.
Log in or sign up for Devpost to join the conversation.