About

Prateek Chhikara is a highly skilled Data Scientist with over two years of experience in advanced analytics, machine learning, and computer vision in the real-estate domain. He is proficient in machine learning and its applications and is adept at solving complex use-cases through data. Prateek is always curious to learn new approaches and tools in Artificial Intelligence and Machine Learning, making him an active bridge between the development team, analytics team, and business. Prateek is a team player who loves harmony and has contributed to the smooth execution of multiple projects at Housing.com. Besides his professional side, he is also passionate about research and writing blogs about Technology. He has actively contributed to the research community since his undergraduate days and has published more than ten research papers in prestigious Q1 journals like IEEE IoT, Elsevier FGCS, and A/A* conferences like IEEE Globecom and ACM Mobicom. At Housing.com, Prateek has made significant contributions, including developing a fraud engine with 92% precision that prevents users from dropping fake leads on the platform and exploiting it. He also devised a computer vision pipeline that creates a 3D floor-plan with the help of equirectangular images and created a system for real-estate entity classification that uses WhatsApp group messages and extracts tags from them with an f1-score of ~91%. Prateek has also automated the manual tagging of images uploaded by the sellers in real-estate classes such as Bedroom, Bathroom, Kitchen, Balcony, Hall, etc., with a recall of 94% and a precision of 89% over all the classes. Prateek's contribution has resulted in savings and an increase in online traffic, and he has been recognized by his company for his efforts. He was honored with the Exemplary Performer for the Year 2020-21 Award and the Spotlight Award twice for each Ownership and Excellence. Prateek's past experience includes working as a Data Scientist at Elara, where he created a model for text detection in an image using VGG-16 with batch normalization as the backbone and achieved a precision of 89% and a recall of 97% using Keras and Tensorflow. Prateek holds a Bachelor of Engineering degree in Computer Engineering from Thapar University. He is proficient in various technologies, including Research Scientist, Data Scientist, ML, Computer Vision, Keras, Jupyter, OpenCV, NLP, Python, Pytorch, Normalization