GeniusZoneQuest

Inspiration

As a working mother of two young children, I know firsthand the constant juggling act that many women face, balancing responsibilities at work and at home. In many households, work is not always shared equally, and the burden can be overwhelming. Often, amidst the busyness, we don’t pause to consider whether the tasks we’re doing bring us joy or negatively impact our health. I’ve experienced that end-of-day feeling—exhausted but still unsatisfied, knowing I gave my all yet still feeling drained.

Recently, I came across the concept of the "Zone of Genius," which identifies the four zones we operate in: incompetence, competence, excellence, and genius. This idea resonated deeply with me. For many women, understanding and identifying tasks that truly energize us remains a blind spot. This inspired me to create an app that empowers women with insights into which activities maximize their energy and which might be draining them. My hope is that this awareness will enable them to make more mindful choices, ultimately leading to a more balanced and fulfilling life.

What it does

The app helps users identify their "zone of genius"—the tasks that energize and align with their strengths. When users input a task and rate their enjoyment level, the app predicts the task’s zone (Incompetence, Competence, Excellence, or Genius) and provides tailored recommendations on how to approach it. By fostering this awareness, the app empowers users to make more intentional decisions about how they spend their time, ultimately supporting a more fulfilling and balanced life.

How I built it

To build the project, I fine-tuned a model in Google AI Studio using Gemini 1.5 Flash, gathering and training data to achieve accurate predictions. I created an interface in Thunkable, leveraging its no-code capabilities, and integrated the Gemini API to enable user interaction with the model. Additionally, I explored Vertex AI and Bubble.io early on to assess various configurations and interface options, ultimately choosing a no-code approach with Thunkable for simplicity and ease of use.

Challenges I ran into

1. As a product manager with limited hands-on experience, the steep learning curve was one of my biggest challenges in understanding and effectively using new tools and technologies.
2. Gathering data and fine-tuning the model to achieve accurate model responses proved to be another major hurdle. It took significant effort to refine the model and train it to deliver consistent, desired results.
3. My initial attempts with Vertex AI proved difficult due to complex configuration settings and limited beginner-friendly documentation, which made it challenging to navigate as a novice.
4. To keep the project no-code and budget-friendly, I needed a free platform that could meet my requirements. Finding one with the right features for my needs added another layer of complexity.

Accomplishments that I'm proud of

Despite having limited hands-on coding experience, I am proud to have independently completed this project. Additionally, I take pride in leveraging AI to address a real-world challenge that many women face, empowering them to make more informed decisions and lead more fulfilling lives.

What I learned

As a product manager with limited hands-on implementation experience, participating in this hackathon has been a valuable learning journey.

1. I gained insight into the time and effort required to explore and select the right tools and technologies for a project.
2. I learned how to use Google AI Studio and fine-tune a model with data tailored to my specific use case.
3. I developed skills in accessing the Gemini APIs.
4. I became proficient in using the no-code platform Thunkable to build the app interface.

Overall, this experience has not only deepened my appreciation for the technology but has also boosted my confidence in collaborating effectively with engineering teams.

What's next for GeniusZoneQuest

Continuous Improvement through Feedback: The next step for GeniusZoneQuest is to implement a feedback loop, enabling users to share insights if the model’s predictions are inaccurate. By capturing their reasons, we can use this data to refine and fine-tune the model, ensuring more accurate and personalized results in the future.

Built With

• gemini1.5flash
• googleaistudio
• thunkable