AI-enabled Analytics for Real Estate Market Research


Infocomm - Data Processing
Infocomm - Natural Language Processing & Semantic Technology


Real estate pricing is extremely unpredictable and complex today, as they no longer follow typical seasonal patterns. Real estate professionals are finding it more difficult to make the right investment decisions and now need to pay close attention to market prices with higher precision and frequency.

Current methods involve manually screening data from multiple listing sites, a practice that is considered outdated and inaccurate.
Such inaccuracies are due to the self-reporting nature of datapoints, e.g. a real estate agent may list a home with inflated sizes to attract buyers. As such, many real estate professionals still rely on manual methods to find the right sale or rental comparables in order to underwrite deals with higher degrees of confidence and precision.

This technology automatically extracts large amounts of data points from both traditional and non-traditional sources to provide accurate and high-quality data. It incorporates Machine Learning, Natural Language and Image Processing techniques to help real estate professionals understand pricing impacts and gain actionable insights.


This technology uses advanced automation tools to monitor real estate pricing trends in granular detail.
Machine Learning techniques, like Natural Language Processing and Image Recognition, are applied to collected data to provide actionable insights for real estate developers and investors.

  • Automatically crawls pricing data at high frequency to detect price movements
  • Data cleansing and standardisation across different sources to provide a uniform schema for subsequent analysis, data fields include (but are not limited to):
    • Unit-type - number of rooms and baths
    • Unit size in square feet (sqft)
    • Unit floor level
    • Balconies
    • Concessions / discounts
    • Price per month (for rentals)
    • Lease terms (for rentals)
  • Uses Image Recognition algorithms to extract salient graphical features from floor plans, in order to determine the nuanced unit types (e.g. urban, penthouse, lofts, townhouse, etc) and flat features (e.g. double vanity, balcony, walk-in-closet, dens/study roomts etc)
  • Uses NLP algorithms to automatically parse texts related to concessions and converts them to dollar amounts
  • The following actionable insights are extracted by analyser modules and presented on a web interface:
    • Price movement for each unit type (monthly rent, rent PSF)
    • Min/Max prices for each unit type
    • Most profitable unit layout and sizes
    • Floor level premiums in high-rises
    • Rental performance of each building
    • Similarity of buildings in terms of amenities and apartment features


Real Estate Developers

Track and underwrite deals with high accuracy and confidence, backed by high-quality, high-frequency data

Property Managers

Adjust pricing based on market changes to maximise revenue

Real Estate Investors 

Enter new markets and fact-check assumptions 10x faster than existing manual methods


Make informed decisions for investment properties and fact check agent/broker supplied information


  • Accurate pricing trends and insights
  • Scrapes multiple data sources, for up-to-date pricing data
  • Market trends and historical data for holistic analysis
  • Ability for users to enter quickly new markets with confidence

The technology owner is looking to collaborate with companies in the real estate industry that can provide in-house data, specifically, accurate transacted prices for home rentals and sales. Additional collaboration opportunities are of interest to the technology owner; specifically, organisations with access to datasets that may have a direct impact on real estate pricing, e.g. images of apartment views, sound pollution maps, as well as property managers, building owners, brokers, and banks with commercial rentals and sales data, which will enable further optimisation of the AI algorithm.

Intelligent Body Pose Tracking for Posture Assessment
Gamified Data Annotation Platform for Supervised Machine Learning
AI-based Optical Character Recognition Engine
Conversation-aware Virtual Patient for Mixed Reality Medical Training
Interactive Plug and Play Software to Embed AI-Powered Mental Fitness Habits in Apps
Synthetically-generated Privacy-preserving Data for Machine Learning
Customisable Sentiment Analysis Tool for Text Data
Chatbot with Global Trade Harmonised System (HS) Product Code Search Engine
Automated Extraction and Processing of Text for Business Insights
High-quality Compliance and Emotion Tracking Analytics for Contact Centre Interactions