You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In-browser Singapore property valuation engine for HDB flats and private condominiums powered by dual XGBoost models (< 1ms JS inference). Features interactive transaction maps and two prediction modes: historical comparable valuation and index-adjusted forward projection.
A complete LangGraph multi-agent system demo using SQL tools, Tavily search, MCP Toolbox, and OpenRouter models — with reproducible notebooks and a full supervisor-led agent workflow.
Autonomous multi-agent system for intelligent HDB resale search — combining geospatial analytics, MRT proximity, and XGBoost price valuation using DeepAgents, LangGraph, FastAPI, and Gradio UI. No database required; reads data directly with SQLite. Dockerized, with reproducible data-ingestion pipelines and support for local LLMs.
🏆 AWS AI Agent Hackathon Entry: Multi-agent HDB advisor using Amazon Bedrock AgentCore, Nova Lite, and intelligent cost optimization. Find your perfect Singapore home with AI.
AI Agent acting as a Singapore HDB buyer's agent assisting in shortlisting, viewing management but also take informed decisions. This is a team of multiple agents working together.
2025 HDB resales, same town and flat type: longer remaining lease is associated with higher price/m² (historical-group medians about +1.20% vs +0.69% per extra lease year). Not a flat's depreciation rate.
Layered SQL pipeline in Snowflake (RAW → STAGING → DATA_MART) over Singapore HDB resale transactions, Jan 2017 to Oct 2026. Power BI dashboard of price per sqm by year, town and flat type.
SG Homie is a data-informed Singapore HDB housing companion combining listing search, seller workflows, structured chatbot assistance, and a research-based CatBoost valuation service built with React, Supabase, FastAPI, and Python.