, a wholly owned subsidiary of HDFC Ltd.
and a digital marketplace for home buying, announced that it has filed for a patent for the first-of-its-kind technology called “Congruence Engine”
. The new real estate congruence system will give property recommendations in real time utilizing advanced machine learning techniques. It can understand home buyers’ unstated preferences by monitoring their browsing patterns on the platform. The patent has cleared the stages of patentability search, prior art search and is currently in the publication phase. The patent has been filed in India and filing in US is being considered.
The system has been engineered to find out the best suitable properties going beyond the stated choices or filter options chosen by the users. It will reduce the search time and dynamically update the results based on users’ interaction with the platform. The engine is currently in the beta phase and will be rolled out soon.
Commenting on this development, Sohel I S, Chief Executive Officer, HDFC RED
said, “With this technology, buyers can expect a more humanized assistance in searching for homes. Till now it was about showing the most relevant results based on the requirements keyed in by the buyers. However, we realized that most people searching for homes are first time buyers and have ambiguous requirements. Hence, we designed the Congruence Engine to read between the lines and understand the buyers’ unspecified preferences.”
The Congruence Engine uses methodologies such as Entropy Maximization, Dimensionality reduction, Moore–Penrose pseudo inverse and Dynamic Weighted K-Nearest Neighbors.