Property platform in Singapore. Agents and buyers want a price estimate for any property before a formal valuation.
We need an ML-powered valuation API: input is a property address + key attributes (size, floor, age, proximity to MRT), output is a price estimate with a confidence interval.
We have 5 years of transaction data, URA (Urban Redevelopment Authority) records, and HDB datasets. The API will be embedded in our listing platform and should respond in under 200ms.
Deliverables: trained model, FastAPI endpoint, deployed on our existing AWS setup, with documentation and a monitoring plan.