TAMA AI

Early stage development

Inspiration

Life is forever fleeting, and can come to and end in an instant. Taking care of your health is something that is so important, yet is severely under-looked still in this day in age.

Change is hard no matter who you are, especially in regards to something as personal and important as your lifestyle and health. Our team wanted to create a simple, yet effective product that can help people introduce healthier and more sustainable practises, without feeling overwhelmed or pressured.

We were deeply inspired by the nostalgic companionship that Tamagotchi's introduced in the late 90's, and wanted to encapture this feeling within our product, whilst also enticing users to become healthier in their day-to-day lives. Similar to the effect that Pokémon-Go had in recent years.

What it does

TAMA AI integrates gamified self-care directly into your everyday routine. You feed your digital cat by snapping photos of your meals, which AI model instantly scores for nutritional value. You log exercise sessions and track your bedtime to keep the cat energised and happy. By mirroring your real-life habits in the cat’s on-screen health, TAMA AI offers playful accountability and positive reinforcement.

How we built it

TAMA AI was built using the following:

Frontend:

React Native
Expo Go
HTML
CSS

Backend:

Python
Astra DB
Torch
Hugging Face
Fast API

Challenges we ran into

As this was the first hackathon for all members of this team, there were a lot of challenges that we ran into, especially early on in the hack. We struggled to narrow down and come up with a concrete idea, and were unsure on how many deliverables we could get done within the time frame.

However, our biggest challenge that we faced was the pivot made on Saturday morning to switch from developing a desktop application to a mobile app (due to grandiose advice from a mentor-thank you Erick :) ). Initially progress was zero to none, since none of us had any prior knowledge or experience in app development. Sadly, the second biggest issue was git commits / merging clashes as well as dependency issues. These two aspects of the hackathon were the most challenging areas to overcome.

Accomplishments that we're proud of

As this was everyone's first time competing in a hackathon, the thing that we are most proud of as a team is actually making it to the end, and having something to show for it. Moreover, we are really proud of diving into a new field of coding and being able to persist and overcome all the challenges that came with it.

What we learned

The biggest things that we learned from this hackathon is:

Time management: Learning to organise our production timeline and set realistic goals under tight deadlines. This was essential to making effective and tangible progress.
Role allocation / team communication: Delegating tasks according to each member’s strengths and maintaining constant communication. This helped us stay aligned and adapt quickly when we pivoted to mobile. -Rapid Prototyping & Adaptability: Switching from desktop to mobile development. This forced us to pick up new frameworks and workflows on the fly, honing our ability to learn and implement new technologies under pressure.

What's next for TAMA AI

We want to introduce a number of new things for TAMA AI:

Social Network: Connect with friends and their pet to make TAMA AI more interactive and fun.
App Widget: Having the option for users to interact with their pet on their home/lock screen.
User Recommendations: Tracking and storing user data in combination with an LSTM model, to produce personalised user prompts/tips.
User & Pet Interaction: Able to 'play' with the pet on the app and/or the widget to bring more emotional connection with the virtual pet.
More complex pet animations: Implementing more complex animation features will enhance user experience.