Inspiration- The pitching session where we heard the poignant story of Dr Swenor's degenerative visual impairment. It inspired us to tackle a practical problem faced by most people about how to approach the new challenges thrown at them by life.
What it does- Allevi-aid is a user friendly recommendation system bringing people who face similar challenges together.
How we built it- We built a recommendation model in Python using an unsupervised machine learning algorithm- K-Means Clustering. The model was integrated with a WebApp using Google Cloud (Firebase), HTML, CSS and JavaScript.
Challenges we ran into- We faced a lot of difficulty in deploying our model to the cloud and integrating the backend (Python model) with the front end User Interface.
Accomplishments that we're proud of- We built a robust recommendation system within the short span of 2 days and made it a full fledged usable product by integrating it with the front end WebApp.
What we learned- How to build unsupervised learning models, Firebase as a database, cloud deployment.
What's next for Allevi-aid- Using speech to text conversion to make the App more accessible. Extracting more features and discovering new patterns to make even better recommendations.
