Inspiration
In the early stages of brainstorming, we initially considered creating a traditional habit tracker—an app designed to remind and motivate users to reinforce good habits. However, as a team driven by innovation and the desire to make a meaningful impact, we shifted our focus. Instead of simply reinforcing good habits, we chose to create an app that empowers people to break their negative cycles, addressing the root cause of their struggles.
Breaking bad habits is often overlooked, even though it can have a transformative effect on personal growth. By focusing on habit-breaking rather than just habit-building, HabitBuster takes a proactive approach to self-improvement. We believe that eliminating negative behaviours not only leads to rapid growth but also creates space for more meaningful habits to develop naturally. With a unique blend of AI-driven guidance and tailored strategies, our app offers users personalized support to conquer their toughest challenges, ultimately setting them on a path toward sustainable, long-term success.
Research has also shown that breaking an existing bad habit is more difficult than forming a new one.
What it does
Even though forming good habits is a path to becoming a better person, bad habits can still drain your mental and emotional energy. HabitBuster is designed to help you break these negative patterns by providing a personalized plan tailored to your specific needs. Our AI generates an actionable daily roadmap to guide you toward your goals. Each day, you'll have the opportunity to reflect on your progress, ensuring you stay aligned with your deeper aspirations. We believe that achieving a goal without understanding its purpose can feel empty, so our journaling feature helps you connect with your inner self, fostering self-awareness and leading to a more fulfilling life.
How we built it
The initial phase began with creating design mock-ups on Figma for some screens of our application. This was immediately followed by the implementation phase, where we communicated using Discord to discuss ideas and technical details.
On the technical side, we all agreed to build a web application using React for the front end and Node.js for the back end. For permanent data storage, we use DynamoDB to store users' sensitive information, including their email, username, password, assessment info, and selected plan for breaking negative cycles. We also utilized existing UI libraries, such as Material UI, to help speed up the front-end development process.
For artificial intelligence, we utilize OpenAI API services, specifically the GPT-4o-mini model, to provide personalized plans based on users' assessment results, as well as their personality, levels of determination, and courage. This information is stored in DynamoDB.
Version control is managed via a team repository on GitHub.
Challenges we ran into
Issue: Lack of Transparency Regarding Sensitive files
Specifically, the secret-data.js file exists in the backend but is ignored by .gitignore. As a result, other developers (at least 2) are unaware of its contents and are unable to run the project locally due to missing configuration or data.
Issue: Inadequate Communication During Mergers
Insufficient communication about ongoing changes across branches has resulted in frequent and complex merge conflicts. Developers are unaware of concurrent changes, making conflict resolution more difficult and time-consuming.
Issue: Committing and pushing code to a collaborative repo was a mess.
Everyone pushed their code to their own branches and merging to the main branch happened at the end, making it difficult for the release manager to merge properly. This could have been resolved if we had followed policies on creating and reviewing pull requests.
Issue: Setting up Generative AI with JSON mode
OpenAI Documentation is somewhat obsolete, making it difficult to initialize an instance of OpenAI API in our project, leading to much time wasted. What is more important is that ChatGPT was of no help, so the only viable solution was to consult traditional StackOverflow.
Accomplishments that we're proud of
The integration of Generative AI:
This is our greatest pride, as we were able to make use of Generative AI and use it for various purposes. It is capable of providing personalized pathways for breaking negative cycles, depending on the user's level of determination, personality and source of motivation.
Nice, beautiful, modern look user interface:
Typography, animation, buttons, and white spacing were carefully chosen to provide the most eye-reliving and comfortable look and feel for our users.
Great collaboration between team members
Our team’s ability to work together, even though we didn’t know each other at first, has been awesome. Everyone’s been super willing to take the lead on tasks and get things done without needing to be told.
What we learned
As the most experienced software developer within the team, I learned the importance of interpersonal communication, and time management as well as how to manage the team, motivate others and resolve complex merge conflicts when they exist. We learned to work effectively with each other despite the differences in our backgrounds and skills
We learned to build a full stack from scratch and to take advantage of the available resources found online.
What's next for HabitBuster
There is lots of room for scalability for our application. This could include
1. Mobile and Progressive Web App Expansion
To make HabitBuster more accessible and convenient, we plan to develop a mobile version of the application. Additionally, creating a Progressive Web App (PWA) will offer users the benefits of both web and native apps, such as offline capabilities and enhanced performance across devices.
2. Gamification with a Point/Reward System and the integration of Community Involvement.
Introducing a point or reward system will add a layer of motivation for users to stick to their habit-breaking plans. Users can earn points for completing daily tasks, journaling, or reflecting on their progress, which can be redeemed for rewards or recognition within the app community. This system will reinforce consistency and encourage users to stay engaged with their goals.
3. Progress Tracker with Cloud Syncing
Implementing a robust progress tracking feature that syncs data across devices using cloud services (Azure, AWS, or Google Cloud) will allow users to monitor their habit-breaking journey seamlessly. With real-time updates, they can view milestones, visualize their progress, and receive insights based on their habits.
4. AI/ML Pipeline for Personalized Plan Refinement
Leveraging an AI/ML pipeline, we aim to continuously improve the personalization of our habit-breaking plans. By analyzing user feedback and self-reflections, our system can fine-tune and optimize the generated plans, providing even more precise recommendations that adapt to the user’s progress and challenges.
5. Dynamic Plan Tailoring Based on Self-Reflection
We plan to enhance the user experience by enabling dynamic plan adjustments based on user self-reflection data. This feature will allow the AI to recommend changing the difficulty level, adding new tasks, or modifying goals to better align with the user's evolving mindset and progress. This will ensure that the habit-breaking process remains flexible and user-centric.
However, scalability is not sufficient. Every technological innovation needs monetary compensation to stay afloat. Therefore, from a business perspective, we need to be marketable.
We will establish and run marketing campaigns to promote our innovative product and explain why it is superior to our competitors, especially in terms of its individualization capabilities.
As part of the marketing campaigns, we will collaborate with media celebrities who are interested in pursuing healthy lifestyles, studying, or working more effectively and productively to advertise our application on their platforms.
Log in or sign up for Devpost to join the conversation.