HeartSpeak

Inspiration

Journalling is a nice way to go back on previous events and keep track of your emotions. However, we faced a problem in journalling - it was too time-consuming and quite hard to remember. This is why we developed HeartSpeak!

What it does

HeartSpeak revolutionises the journaling experience. It listens to your conversations throughout the day, keeping track of your heart rate data. At the end of the day, it will tell you moments through your day when you experienced an elevated heartbeat, and based on the sentiment of your conversations, will tell you if it was a positive or negative moment. Additionally, HeartSpeak provides you with various prompts to help you make the most out of the journalling experience, and truly reflect on the various magical moments you experience throughout the day.

How we built it

Our team has a range of skills from frontend to backend and UI design. We split up into teams and pair-programmed. For the front-end, we used ReactJS, TailwindCSS. For the back-end, we developed APIs using Python Flask, managed requests and inputs using Next.JS. We used Figma for UI design.

Challenges we ran into

We ran into problems with inaccurate ML models for speech-to-text. Another challenge was integrating Flask with Next.JS stack, and also collecting heart rate data from our wearable devices.

Accomplishments that we're proud of

Real-time heart rate tracking
Natural language generation
Human centred design

What we learned

We improved our coding skills, especially brushing up on our ReactJS skills
We learned the importance of choosing a good idea early
We learned the importance of teamwork and task decomposition for each sub-team.

What's next for HeartSpeak

Currently, we only classify moments as positive or negative. In the future, we would love to provide support for more fine-grained emotions, such as anger, and excitement. We would also like to track things other than heart rate, such as stress levels and sleep patterns, to give a more holistic view of the user's day.