HumanCell

App that serves to map the energy of the user so they can improve energy management for the day, and perhaps schedule their consecutive days or weeks to complete tasks faster and more efficiently.

Comment

Inspiration

For the past month, a team member(Lok Sang) was managing his daily energy fluctuations, and was experimenting on optimizing the duration and timing of periods where his energy was high enough to perform productive tasks. Thus we thought that an app that visually displays the user's estimated mental and physical energy levels would immensely help the user craft and refine their schedules accordingly to effectively get more work done and leave some truly guilt-free personal time by the end of each day.

What it does

HumanCell's core functionality is to provide an easy-to-understand approximation of the user's energy levels given details on the user's physical attributes (like age) and habits (nature of work, hours of sleep). The approximation can change continuously when fed information regarding the activities of the user. The intended purpose is for the user to be aware of how different activities realistically affect their total mental and physical energy levels throughout the day and for them plan their daily activities around it, and/or actively work to improve their energy levels. Too often, people may think that they are chronically tired or unproductive due to time constraints or character flaws, when in many cases it is just a matter of when you do what, and the nature of your habits.

How we built it

When first run, HumanCell has the user enter a small number of personal details initially to not only profile them and save their data to the cloud for future modifications, but to also provide an initial estimate so the user can immediately make use of HumanCell as quickly as possible. The user can now input more details of their activities continuously anytime they wish to get a more realistic estimate for their energy levels at that time. For instance, the user could input 'I took a nap for 20 minutes', and energy levels will increase; in contrast, 'I took a nap for 15 hours' will most likely decrease energy levels. The evaluation of the initial input on the first page is carried out by Gemini's AI and it returns the starting 2 bars of energy levels. The subsequent, continuous, evaluations actually does not require generative AI to give or reduce points from the energy levels, and is in essence a large function.

Challenges we ran into

We did not have the extensive knowledge required of TypeScript, Python, JavaScript, React.js nor knowledge of the necessary libraries to carry out this project on our own, so programming-wise we used generative AI to aid us in making the majority of this program. We had a lot of trouble trying to fit our different components together and getting the AI to do what we wanted, so part of us wishes we already had that knowledge since it would have taken a similar amount of time, and we would have more control over the project. There is already a feature where we use Gemini to generate suggestions on what to do next given the user's current energy levels, but we had trouble connecting it to the frontend for display so it is currently unused in this version. Given more time we surely would have completed this part.

Accomplishments that we're proud of

We are glad the cloud and AI tokens were integrated quite seamlessly into out application in its current state. The fact that we implemented 80% of our initial planning and got the app to work is an accomplishment in of itself. We also love how the app looks as well, given its classic monopoly theme.

What we learned

We learned how to integrate AI Tokens into our app, as well as connecting the frontend and cloud database to the backend. We also had learnt the importance and usage of APIs. Importantly, we learned how to use generative AI to our advantage when trying to bring a product to launch in under 24 hours. Last but certainly not least, we learned how to collaborate effectively using git commits and branches other than main.

What's next for HumanCell

Firstly we would love to implement the suggestions chatbot feature, as that would really be helpful for the user's next plannings. If possible, more scientific/accurate ways to get the user data or energy measurements will be implemented in the future. We may add multiple tabs for expanded functionalities. Additionally, more parameters will be considered in all evaluations. We will work on this app over the winter.

Share this project:

Updates