BestieBoost

Inspiration

Every year we hear people say "new year, new me" but statistically after 3 months, 90% of these people will quit going to the gym. Although there are so many fitness apps, we wanted one that helps beginners find the right realistic goals, help them get started and keep them accountable.

What it does

We have 3 main functions to our app. Firstly, we have our goal setting page, as people who write down their goals are 42 percent more likely to achieve them than those who don't. As beginners it is often difficult to set realistic goals or feel motivated to workout for the long run which is why we created our second function, the "Bestie Chat" where users are able to chat with their "Besties". They are able to ask questions that help them better set their goals and keep them motivated! Lastly, our main page will be a streak keeping page that tracks the user's progress. It will be a fun and interactive log where every time they want to log themselves working towards a goal, they can show themselves in the action and our computer vision technology will classify whether or not the goal was completed. They can have fun in the moment but also keep long lasting memories of their progress day to day. It will act as a little reward for them even if they can't see the physical changes yet, they are able to see their hard work and effort.

How we built it

Our main function and the app layout was created using typescript. In order to use computer vision to scan our object we created a javascript file and used ml5 to generate the video and classification. For the "Bestie Chat" we used DeepChat for the core structure and made it functional by connecting it to Cohere's API. The "Bestie Goals" was also made in typescript. The login was created using Clerk Authentication which connects it to Google's login system.

Challenges we ran into

We had one dev working on the computer vision portion of our app and another dev working on the rest of the structure so when we had to link the two it was very difficult. Originally we were unsure if the computer vision would work so we didn't want to take any chances so one dev worked on a different repository all together and when she finally got it to work we tried to move all the files into the rest of the code. In the end we found a way to reroute the originally page to the other page that the computer vision was stored in.

Accomplishments that we're proud of

We poured our hearts into crafting Bestie Boost's technology. Our user interface is just the tip of the iceberg – the underlying tech stack is carefully chosen to address the common challenges users face when starting their health journeys. We tackled user intimidation head-on, prioritizing features that make getting healthy feel approachable and engaging. A particular highlight was implementing computer vision using ml5, a new technology for our team. Seeing it come to life within Bestie Boost was incredibly rewarding. We also had a blast creating the cast of characters (Lebron, Aqua, Swol, and Paul) for Bestie Boost, injecting a touch of fun into the development process.

What we learned

Bestie Boost was a valuable learning experience that honed our product development skills. We delved into market research, understanding the importance of considering every facet of a product – from user needs and feature implementation to monetization strategies. This experience solidified the concept of the Minimal Viable Product (MVP) for us, a valuable takeaway for future endeavors. Bestie Boost also instilled a growth mindset – we conquered the challenge of implementing ML5 image classification, proving our ability to learn new technologies. Each team member emerged with a deeper understanding of both the business and technical sides of product creation.

What's next for BestieBoost

To boost long-term engagement, we're considering integrating a character leveling and customization system into BestieBoost. This gamified approach would tap into users' intrinsic desire for progress and achievement, motivating them to complete their wellness tasks and stay invested in their health journey. By allowing users to personalize their characters, we can foster a deeper connection with the app, making healthy habits more enjoyable and sustainable.