OpenAI AI systems are transforming Heimstaden’s property management
Heimstaden is using OpenAI tools to transform the management of thousands of homes, from listing properties to assessing environmental risks. In the long term, the company envisions properties whose operations are supported by increasingly intelligent systems.
- Heimstaden chose OpenAI in June 2025 to accelerate its AI strategy.
- In the Netherlands, the cost of producing virtually staged images fell from 50 euros to less than 0.10 euros.
- Using Codex, the company assessed the environmental risks of 20 112 properties.
- ChatGPT Enterprise reached 50 percent of employees within two to three months.
- The company is preparing AI-powered workflows for its property management systems.
AI supports the entire rental process
Heimstaden’s leadership decided in June 2025 to accelerate the use of AI. After evaluating several platforms, the company chose OpenAI. The aim is both to support employees and rethink business processes.
One of the most important areas is re-letting vacant apartments. AI generates rental price recommendations based on ChatGPT Sites tools, produces virtually staged images and property descriptions, and checks active listings against Heimstaden’s quality standards. The system also generates market feedback that rental teams can use to adjust prices.
Employees remain responsible for reviewing and approving decisions. In the Netherlands, the cost of producing virtually staged property images fell from around 50 euros to less than 0.10 euros. Heimstaden estimates that this alone represents an opportunity to save 130,000 euros annually in the country.
According to Helge Krogsbøl, co-CEO of Heimstaden, the purpose of the buildings has not changed: they continue to provide homes for people. However, the way the company operates is constantly evolving, and Heimstaden sees AI as one of the biggest opportunities to accelerate this transformation.
Codex enables specialists to develop solutions
The change is not limited to traditional software development teams. Heimstaden wanted to assess environmental risks across its property portfolio, but the necessary information had to be collected from multiple public sources, in different formats and datasets.
Julia Erlandsen, the company’s environmental sustainability analyst, developed a new approach with the help of Codex despite having no programming background. The tool supported data discovery, cleaning and standardisation, quality control, and the creation of repeatable processing workflows. Erlandsen directed the methodology and review process.
During the project, 20 112 properties were assessed, generating more than six million rows of useful data. The estimated workload fell from one person-year of work by five people to two months of work by two people. Heimstaden also avoided bringing in external consultants. According to the company, the solution could support climate risk assessments, reporting obligations and investment decisions.
Training and central governance accompany adoption
Heimstaden’s AI strategy is built on three pillars. The company supports employees through training and appropriate tools, operates a central transformation office and AI ambassador network, and prepares its property management systems and data models for AI use.
After the launch of ChatGPT Enterprise in September 2025, half of the employees were using the system within two to three months. The central AI transformation office maintains a library of use cases so colleagues can learn from examples within the company. Employees use approved tools under data security rules, while reviewing AI-generated content and approving decisions remain their responsibility.
Adoption also extends to property managers. In Norway, a caretaker used ChatGPT to assess a neglected apartment and then created a prioritised renovation plan that could later be reviewed and refined.
The next step is AI-supported action
Heimstaden is already working on systems capable of coordinating tasks across multiple applications and workflows. The Procurement Intelligence Platform, currently being introduced, links workflow automation with models accessible through OpenAI APIs. It was designed for use with Outlook, SharePoint and Teams, and supports contract analysis, negotiation preparation and management of the full contract lifecycle. Commercial decisions and approvals remain with employees.
The company is also working on Amplio Estate, a jointly developed property management system. The system is intended to support AI-powered workflows through MCP, enabling employees, for example with the help of ChatGPT, to retrieve property data and perform actions in core business systems under appropriate human supervision.
Heimstaden’s vision is not of buildings without people. The company wants properties supported by more intelligent systems that improve employees’ decision-making and their ability to resolve customer requests.
