Hi Garmi

Inspiration

Sitting across from each other, we realized our countries' shared struggles with increasing temperatures and the total lack of infrastructure to face them. We also saw how important community wisdom is in tackling this issue and we wanted to tap into that in our own way, both to highlight the problem and to provide solutions. Since climate change affects everyone, everyone should have a seat at the table with a healthy exchange of local solutions and ideas to make global progress. The SDGs adopted by the UN served as inspiration too, with us being based in the international organization hub Geneva. Specifically, our work falls under SDGs 3 - Good Health and Well being, 11 - Sustainable Cities and Communities and 13 - Climate Action

What it does

Hi Garmi! - Inspired by the commonly used South Asian phrase “Haye garmi!” expressing the frustration at how hot it is. The project is aimed at addressing the urban heat island effect, where urban areas experience significantly higher temperatures than surrounding rural areas. The primary goal is to create a crowdsourced platform to collect, analyze, and respond to data concerning the urban heat effects on both individual and governmental levels. We want to turn the general sentiment around so that people are ready to meet and beat the heat.

How we built it

• Step 0: Research and Data Review We looked through various sources of data on climate change, urban heat island effect and realised the need for more granular and localized data collection

• Step 1: Data Collection We created a survey to know about people's experience with the heat, the personal measures and habits they adopt to deal with high temperatures and their suggestions to the government bodies. We based our surveys on urban heat mitigation plans made by cities across South Asia. We aimed for representation of different cities and backgrounds as far as possible.

• Step 2: Data Analysis and Visualization We created a Streamlit dashboard with word clouds and graphs and we saw patterns and aggregations emerge in the data showing concretely the similarities of the effect of climate change among different cities and countries. We realized we could use this data to create a chatbot for easy and efficient querying of the collected data.

• Step 3: Building a Chatbot We created a chatbot using Chatbase with the data collected from the respondents. We were able to use it to generate comprehensive, quality responses on area-based personal and policy measures to combat urban heat island effect.

Challenges we ran into

• User-friendly survey design: Building a survey that captured all the data we needed and yet, was not cumbersome.
• Crowdsourcing Data: Reaching out to people and building trust in a cross-country collaboration

Accomplishments that we're proud of

• We were able to reach out to communities and collect data to build our RAG on real data from real users
• Although we started with South Asia in mind, we have built a blueprint for a forum for discussion of heat-mitigation strategies that can be applied to any city across the world.

What we learned

• Complexity of the problem and how badly people are looking for a solution
• Building with Chatbase and Streamlit
• Creating visualizations to communicate patterns in opinions

What's next for Hi Garmi

• We hope to collaborate with international organizations working in these fields like the WMO, WHO etc.
• We want to build an end-to-end solution uniting the approaches that we took to inform policy discussions on climate change and urban heat island effect. We want to become the go-to platform for this issue.

Built With

• chatbase
• gpt
• llm
• python
• streamlit