Inspiration
Drawing inspiration from industry giants like Amazon, we recognize the power of advanced algorithms in driving personalized recommendations. However, such resources are often out of reach for small businesses. Clusteroo bridges this gap, empowering small e-commerce ventures with data-driven insights and recommendation capabilities previously accessible only to large enterprises.
What it does
Clusteroo provides each client with a custom machine learning model to enhance their ability to target customers with tailored product recommendations and marketing strategies. We call these models Joeys.
How we built it
We built Clusteroo with a comprehensive tech stack, incorporating Flask for the backend API, MongoDB for data storage, and Twilio and SendGrid for communication services. The frontend was developed using React, specifically Next.js, to provide a seamless and responsive user interface. This combination allowed us to create a powerful platform that enables users to upload CSV data, train machine learning models, and receive tailored recommendations, all through an intuitive and engaging interface. Additionally, we utilized GridFS to efficiently store large files within MongoDB, ensuring scalability and performance.
Challenges we ran into
The big challenges were mainly focused around the integration between our front end and back end. We had to ensure a smooth user experience with our APIs.
Accomplishments that we're proud of
We have our first client and will be starting business with them next week!
What we learned
We refined our skills in the coding languages we used while also learning some new frameworks like scikit-learn. We were able to take advantage of the knowledge we had individually to teach the other as we developed the project. This was our first project implementing Twilio and SendGrid.
What's next for Clusteroo
We will gather data on businesses, cluster them, and target our own marketing campaigns towards the businesses that align with our target audience. We will then create an AI meant to fine tune our customer's models.
Log in or sign up for Devpost to join the conversation.