BrainDump

Home Page
Input Fields
Personal Project Roadmap

Inspiration

Have you ever wanted to learn a new skill but had no idea what to do or how to start? All members on our team agree that they have faced this issue multiple times. What students like us need is a medium that takes our ideation process to the next step: research and implementation. But research is laborious and requires a significant dedication of our time, making students like us forget about our amazing ideas sometimes!

What we look for is a software tool that enhances the creative products of our thinking with the vast amount of knowledge and information present on the Internet. We look for a tool that can link our various ideas and learning into personal project ideas that can be implemented with clarity.

What we look for is BrainDump, an AI-powered web solution that transforms our elementary ideas into impactful personal projects that will provide great experience to our users. With BrainDump, we hold access to a powerful asset that eliminates the issue of idea-to-implementation stagnation and empowers our users to start working on what they enjoy.

What it does

Our website asks the user for what they would like to learn in the form of a personal project. After the user submits their interest, our Google Gemini-backend produces a clear roadmap on what to learn/do to complete the project. Specifically, our website displays the project title, a succinct description, and a list of “stages” that can be followed by the user. These stages include the stage name and a description of the stage and the skills it requires.

On a side note, if the user’s browser is kept open, then the user’s session is saved, meaning that the same roadmap can be accessed again.

How we built it

We had four group members, so we divided and conquered. Geethan and Raheel worked on designing the front end first in the Figma design software, then implementing them in code using HTML/CSS and JavaScript. Zaid was in charge of creating the custom AI pipeline using Google's Gemini API. This API allows the user to submit ideas that are then processed by the Large Language Model. The AI outputs a personal project roadmap which includes a description and stages of completion. Arjun was in charge of linking the front and back ends together using the Flask web framework, allowing the user inputs to talk to the AI and allowing the AI outputs to be displayed on the front end. Throughout the development process, best practices were utilized to ensure proper documentation of our codebase through GitHub.

Challenges we ran into

There were many challenges we ran into during our programming process. One such challenge was the countless merge conflicts in Git and debating over which lines should be incorporated and where. Other challenges include bypassing Google Gemini's content filter so that random errors would not occur because the API deemed it too unsafe. On the front end side, there were many difficulties with formatting certain tricky areas and creating the vision designed on the Figma software.

Accomplishments that we're proud of

We were mainly proud of our success in combining all elements of the project together. When we started, we were unsure if the frontend, backend, and API would actually end up working together. While each element of the project was a grand development for us, seeing the elements all work together was truly rewarding in the end.

For those of us who this was the first hackathon of, we were especially proud of working under the time crunch to push out a project that we are satisfied with.

What we learned

We learned many new skills during this competition, including API integration, Flask, GitHub, and more to structure our program around. For half of us, this was our first hackathon, so to be able to perform as well as we did is amazing. However, we do need to plan our time better and learn more frameworks so that we are able to deploy them much faster. That way, we can create a more complex and fleshed out product than what we were able to accomplish this time.

What's next for BrainDump

There are many features we plan to implement, including dynamic dates that change based on whether the user hits a deadline or not, animations that reward the user if they complete a task before the deadline, a user authentication system to support multiple users, auto-generated links to resources that could help the user to complete their personal project, the ability to track multiple project ideas, faster response times, and more.