Inspiration

Singapore publishes some of the richest housing data in the world, yet young buyers including us, still have no clear way to see what they can afford or where resale prices are heading. Resale prices have risen sharply over the past decade, and the 2024 flat classification has reshaped the market again.

In 2026 Q1, the number of resale applicants reached 12,676 (HDB, data.gov.sg), suggesting more buyers are entering the market now. We wanted to turn open data into something a first-time buyer can use before committing to a long-term purchase.

What it does

Resale Therapy is a dashboard built on HDB resale data. It has three features:

Affordability check: users enter their income and the dashboard shows whether flat types in specific towns fit their budget.
Price forecast: a 6-month price outlook for that particular town & flat type with an uncertainty range.

How we are going to build it

We used HDB resale data and other data from data.gov.sg. We cleaned and loaded the data, built the trend views, implemented the affordability logic, and connected everything in one interface.

Databricks components we will be using include Lakeflow, Delta Lake, Unity Catalog, Databricks SQL, Genie, MLflow, Databricks Apps, AI/BI Dashboard.

Challenges we expect

Data cleaning: in the event of inconsistent flat type labels and how changes in data is recorded over time.
Two-week scope: we plan to keep the forecast simple for predicting of resale prices within the next 6 months so that we can deliver a working product.

What we hope to accomplish

A working end-to-end dashboard built within a two-week sprint
A forecast that shows its own uncertainty, so users can judge how much to trust it

What's next for Resale Therapy

Add a short feedback survey to measure whether users feel clearer about what they can afford
Test more advanced forecasting methods against the baseline
Track weekly usage to see whether adoption is growing