Kora | Devpost

Kora Landing page
Kora user personalization
Kora Mobile app

Inspiration

In a digital landscape overflowing with information about climate change and our carbon footprints, it's easy for users to become overwhelmed and uncertain about their actions. Our project was inspired by the desire to keep our users informed, simplifying often complex decisions that we simply don't have time to consider with our busy lives.

What it does

Kora is a travel planner that calculates the carbon footprints of your journey to help you make an informed choice about your travel routes and methods. Its Agentic AI helper allows you to choose your destination country, which cities you want to go to, and even recommends potential landmarks you could visit! All the while providing insight into the impact of your trip on the planet.

How we built it

We designed the UI in Figma and extracted the designs into Tailwind (JSX) format. To project our vision of KORA, Next.js and React are the front-end frameworks used alongside TypeScript, TailwindCSS, and HTML. Moreover, npm libraries were used in visualizing the 3D object and making it interactive.

With a Python-based back-end, we are leveraging the Flask library. Our agents are based on the Gemini AI API. In collaborating and connecting our AI Agents, Langchain was used to chain Agentic AI. Secure authentication is ensured by using auth0, a profile authentication API that allows users to sign up and log in to KORA. Users' data are stored using the MySQLite database management system.

Challenges we ran into

The biggest challenge of implementing our solution was creating a functional and accessible user interface that would welcome users instead of overwhelming them. Discussions surrounding carbon footprints can often feel intimidating and daunting. Our goal is to reframe that negative experience into a rewarding one—making sustainability feel fun, not scary. The finalised design of Kora as our mascot symbolises the project’s environmental and oceanic inspiration through positive semiotic cues.

Accomplishments that we're proud of

We are proud of our idea, which promotes sustainability while keeping things simple, instead of overwhelming. Moreover, the eagerness to learn and create in the team is what drives and fuels us to integrate improvements to KORA continuously.

What we learned

Implementing the LangChain Agent added a layer of functionality and personalisation to our end project. However, we also learned that it is important to give strict prompting instructions to LangChain to prevent it from giving unclear output. Additionally, we got practical experience in exploring tools and approaches by creating KORA, such as using AI Agents and libraries for the base of the application.

What's next for Kora

In the future, we plan to introduce a friends feature that includes a leaderboard to track users’ progress. By fostering friendly competition and social accountability, this feature could encourage greater mindfulness about one’s carbon footprint. The hope is that our project could potentially inspire long-term habit changes and have broader, positive impacts on society. Lastly, we see that there is a lot of room for future improvements to create a better usage flow and increase users' experience.