Introduction: We are the HackerBees a team of five software engineers with an average experince of one year in coding.
Framing the Problem: The team experienced significant delays in settling on a final idea during the hackathon, taking nearly a full day to reach consensus. Attempting to utilize the Hackathon Learning Centre for guidance proved ineffective, we are highlighting the need for a more efficient ideation tool that not only enhances team productivity but most importantly improves the ideation process by taking advantage of AI.
Idea Explanation: Our solution incorporates AI to help teams analyze their ideas, find the common themes and prioritise those that are more likely to generate the best solution to the problem.
- When logged into a team’s workspace, the problem of the challenge should be added under the challenge name (that's important for later).
- By registering your team’s interest, the team’s getting the pop up with the notification that a new task has been added into the tasks. The "Decide on the idea" task has the link to the Rapid Ideation tool.
- The Radip Ideation tool has different idea generation techniques that a team can use: crazy 8, brainwriting, mindmapping, etc…
- Crazy 8 demo: each room loads for every team member, and by clicking “I’m ready to generate ideas!”, the clock starts. You’re presented with the timer (8 min), the input area for the text, the blocks that each idea will occupy, and the problem on the top. If the team member really struggles (after 2 min or so - no ideas submitted), the questions start popping up to help them with idea generation. After the timer runs out, all ideas get fed into AI, that does the magic: It groups all team member’s ideas to find the common themes and analyzes those groups of tasks from best tackling problem possibility to least. The AI gets the problem prompt from the record of registered interest on the Challenge registration page. Then AI outputs results for an open discussion, suggesting running the discussion against 3 rules: simplicity, relational and potential. Result: clarity about the ideas.
Tech Stack: How do all the pieces fit together? The tech works with python and html pages. Python is used to initiate a localhost to view the html pages. Additionally we have used Google Gemini AI API. The api key has to be already hardcoded into the terminal as environment variable. Does your frontend make requests to your backend? Yes, it makes post and get requests for multiple things to the python script. The python script takes get requests for html pages and posts the team-members idea pages and to the gemini ai. Where does your database fit in? We don't use a database but use json format to parse data from the Gemini AI. Give a high-level overview of what SOFTWARE come together to create your project??? We use HTML, CSS in the frontend and Python as backend.
Technical Deep-Dive: The python script when run initiates a localhost using HTTPhandler library and renders pages according to requests. When the ideas are generated then it is sent to the Google Gemini AI API and it summarizes based on all the ideas and then returns back all the data. This is then parsed into taking the important text only and then it is rendered.
Challenges Faced:
- Time Constraints
- Ideating: Ideation process took more time and this caused us to start building later.
Accomplishments Generated a working tool that satisfies the desired outcome - to make ideas simpler. Great team work under the time pressure - gain invaluable collaboration skills. Utilized to the best each others strenghts.
What are you proud of?
- Finally coming up with an idea for us and for future atendees experiencing the same problem, and making Hackathon's a better experience in general.
Next Steps: Potential features include -Add a fun elemement at the end of the ideation phase - GetToKnow -Add the information about uniqueness of your solution against other people’s solutions. -Value for stakeholders: Access to all the solution ideas to see if there’s anything interesting.
Project Links: Presentation Link: "https://www.canva.com/design/DAGUwY_R-40/RpgxVz2AxmxCiyaA9FQYpg/edit?utm_content=DAGUwY_R-40&utm_campaign=designshare&utm_medium=link2&utm_source=sharebutton" Github : "https://github.com/moulish-dev/HackerBees"