KopiTalk

Inspiration

We want to help volunteers identify neighbourhoods where elderly people may be at higher risk of social isolation. KopiTalk will use public data to make outreach more targeted and effective.

What it does

KopiTalk will identify high-risk neighbourhoods, find gaps in nearby support services, and help volunteers decide where to go and how to start a conversation.

How we built it

We plan to build KopiTalk using Databricks and Singapore's public datasets. ML will analyze factors such as elderly population, housing, healthcare access, and nearby services to generate neighbourhood-level risk scores.

Challenges we ran into

The main challenge will be combining different public datasets and creating a meaningful risk score without identifying or predicting risk for individual people.

Accomplishments that we're proud of

We are proud of designing a solution that goes beyond simply showing data on a map. KopiTalk will connect risk detection → service gaps → volunteer action.

What we learned

We learned how public data can be used to understand social problems and how explainable ML can support better community decisions.

What's next for KopiTalk

If shortlisted, we will build the working prototype in Databricks, test the risk predictions, create the dashboard and outreach assistant, and improve the system using evaluation results.