Inglatan streamlined workflow of user-uploaded property image compliance and categorization of photos with AI photo recognition
Nearly 20 years old, ingatlan.com is one of the oldest Hungarian real estate websites. They are proud that the company has been in Hungarian hands since the very beginning with their founder still being the strategic leader of the company as it’s grown to more than 120 employees. ingatlan.com is committed to understanding the real estate market’s needs and is working to provide world-class service to their customers.
Problem
As a leading real-estate portal in Hungary, ingatlan.com always strive for the best. Their purpose is to gather all real estate listings in one place, while maintaining a complete, ethical, and effective real estate market in Hungary. After more than 10 years, the platform now organizes professional events and publishes regular printed publications.
It’s clear that ingatlan.com has an impressive and busy record in the property market. As a Product Architect overseeing content quality for the company, Péter Szabó realized that ingatlan.com needed some help in moderating and controlling the bad content that more and more users uploaded on the portal. With more than 30,000 pictures uploaded daily, manual screening would be an expensive function. Therefore, Péter started looking for a more sustainable solution to tag and verify their images.
Solution
Many real estate portals and agencies prioritize self-branding over user experience; ingatlan.com often saw that these agencies put any type of contact information (from phone numbers to logos or text) in the center of the uploaded images. Péter needed a solution that could automatically detect and tag this type of information on images, to enable him to moderate the content and improve ingatlan.com’s user experience and engagement.
Understanding that developing a machine learning solution in-house would be costly and time-consuming, ingatlan.com decided to opt for a modern, automatic tool, instead of hiring 5 new content moderators. Moreover, they wanted to stay focused on their core competency: the real estate market. Comparing Google Vision’s API, Amazon Rekognition, and Clarifai, ingatlan.com found those solutions were not suitable for its needs. After much investigation, they chose a company with expertise in computer vision tailored for the real estate industry: Restb.ai.
ingatlan.com successfully integrated Restb.ai’s computer vision API, reaching the best workflow for their users. They were very satisfied with the speed and accuracy of the results and have been using it since the middle of 2017. Given that Restb.ai is the leading company offering image recognition solutions for real estate, it provided the expertise that ingatlan.com needed.
Impact
Before integrating the image recognition API to detect, tag, and categorize photos and text in images, ingatlan.com struggled with quality control of more than 30% of the images uploaded on their platform daily.
However, thanks to the incredible capacity of Restb.ai’s solution to flag logos, watermarks, and any type of text on property images, ingatlan.com’s website is now fully optimized and visually appealing. Users cannot upload messy, low quality and self-promotional material anymore, but clean, high-quality pictures, with no irrelevant text.
Since then, the amount of clear and high-quality pictures on the platform increased significantly. Péter can now rest assured that even if individuals try to upload inappropriate content, the API will automatically detect it and the process will be prohibited.
ingatlan.com’s portal remains clean and provides an excellent user experience. Visitors become more engaged and spend more time on the website. Ultimately, ingatlan.com enjoys improved lead generation.