AI-enhanced photo editing in real estate marketing

A case study of Esoft's adoption of a Hybrid Intelligence framework

This project is a collaboration with Esoft, a multinational company specializing in the acquisition and editing of real estate images. The core task at Esoft involves transforming multiple exposures of real estate properties into a single corrected image. This process is carried out overnight by an overseas image manipulation unit in Vietnam. Esoft serves as a prime case study due to its dynamic integration of AI and human expertise. By embedding AI in image processing for photo editors and enhancing the capabilities of sales professionals with data analytics, Esoft demonstrates the substantial potential of human-AI collaboration.

The project began with the establishment of the 'Development Life-Cycle for Hybrid Intelligence Systems,' a framework that guides the design of Hybrid Intelligence (HI) applications and spurs further research in HI design and implementation. This framework was applied to enhance Esoft's industrial photo editing processes, leading to the development of the 'AI+Human Photo Editing Web-Application' prototype.

The prototype integrates interactive machine learning to foster collaboration and mutual learning between humans and AI. The focus then shifted to a 'Strategic AI Transformation Analysis' within the real estate marketing industry, exploring the potential for AI to disrupt the market. Through extensive interviews with industry leaders, we assessed the potential of AI-driven transformations and proposed HI business model innovation solutions for the company.

Based on the premise that generative AI and interactive machine learning could free up valuable employee resources, we interviewed sales staff about product features they would love to market but previously lacked the organizational capacity to sell.

A dynamic model of human-AI collaboration

An example of HI-driven value generation innovation

An example of an emerging, HI-driven innovation in mass customization and value generation is the proposal to develop a plus-service. In this service, homeowners could express one or more emotions about their house.

These emotions would then be relayed to the image processing unit in Vietnam, allowing photo editors to shift from routine tasks to specialized services—transforming their role from production line work to creative studio activities.


Christian Bantle

Former Research Intern

Jacob Sherson

Our funders