How Heimstaden Deployed OpenAI to Cut Property Image Costs by 99% and Automate Letting Workflows
European residential real estate operator Heimstaden has integrated OpenAI’s ChatGPT Enterprise and Codex into its property management operations, reducing the cost of producing virtually staged property images in the Netherlands by more than 99%. Announced on November 3, 2026, the deployment shifts the company’s approach to letting vacant homes and assessing environmental risks, replacing manual coordination with AI-assisted workflows across its portfolio of 158,000 homes.
How ChatGPT Enterprise Restructures the Residential Letting Workflow
When a residential property becomes vacant, traditional management requires assessing rental value, arranging photography, producing advertisements, and responding to market demand. Each stage historically demanded significant time, coordination, and specialist input. Heimstaden is now using AI to support every key step in this process.
According to unite.ai, “AI now supports pricing recommendations through tools built in ChatGPT Sites”, styles property images, and writes apartment descriptions. An AI agent reviews all live advertisements against the company’s quality standards, while AI-generated market feedback helps letting teams take corrective action to make pricing more dynamic.
Employees remain responsible for reviewing and approving final decisions, but the analytical and production work surrounding those decisions is automated. In the Netherlands, the cost of producing a virtually staged property image fell from approximately €50 to less than €0.10. Heimstaden estimates this yields an annual savings potential of €130,000 in that market alone.
Nikolaj Brammer, Heimstaden’s Chief Digital Officer, explained the methodology: "We started by breaking down our most important business processes and asking where AI could create value. Letting is a good example: it’s not about automating one task, but improving how the whole process works."
How OpenAI Codex Enabled a Non-Technical Analyst to Build Data Pipelines
Beyond the letting workflow, Heimstaden needed to assess environmental risks across its vast property portfolio. The necessary information existed but was scattered across public sources, formats, and datasets. Turning this fragmented data into a usable assessment traditionally required substantial manual work and dedicated software engineering resources.
Julia Erlandsen, an environmental sustainability analyst without a software development background, used OpenAI’s Codex to build a new approach. Codex supported the discovery, cleaning, and standardisation of data, quality checking, and the development of repeatable processing pipelines. Erlandsen retained control over the methodology and validation throughout the project.
The results were substantial. The initiative assessed 20,112 assets and generated more than six million rows of useful data. This reduced the estimated delivery requirement from five people working for a year to two people working for two months, while also allowing the company to avoid engaging external consultants. What began as a complex analytical project is now a repeatable capability supporting climate-risk assessments and investor decision-making.
Why Heimstaden Accelerated Its Enterprise AI Strategy Across 158,000 Homes
The timing of this deployment stems from the sheer scale of Heimstaden’s operations. The company manages a portfolio of approximately 158,000 homes valued at around €30 billion across nine countries, including Sweden, Germany, the UK, and Poland. Maintaining operational efficiency at this scale without proportionally increasing headcount requires moving beyond isolated AI experiments to enterprise-wide integration.
In June 2025, Heimstaden’s leadership evaluated several platforms and chose OpenAI to accelerate its AI strategy. The approach is built on three pillars: empowering employees through training and tool access, establishing a central transformation office supported by a network of AI champions, and strengthening foundational property management systems and data models.
Helge Krogsbøl, Co-CEO at Heimstaden, summarized the strategic intent:
The purpose of our buildings hasn’t changed. We provide homes for people. But the way we operate them continues to change as we challenge conventions and adopt new technology. — Helge Krogsbøl, Co-CEO, Heimstaden
What IT Decision-Makers Can Learn from Heimstaden’s AI Deployment
For IT leaders and property managers evaluating AI tools, Heimstaden’s rollout offers concrete lessons on enterprise AI adoption. As noted by welcome.ai, the company achieved significant operational savings and a more responsive letting process by treating AI as a process redesign tool rather than a standalone utility.
The most critical takeaway is that AI efficacy is strictly limited by underlying data readiness. Heimstaden’s third strategic pillar—making property data accessible to AI tools—was a prerequisite for the dynamic pricing and automated listing creation seen in the letting workflow. Without clean, accessible property data models, the AI agents could not execute quality reviews or generate accurate market feedback.
Furthermore, the deployment highlights a shift in how businesses approach technical development. By utilizing Codex, Heimstaden enabled a domain expert to build a complex data pipeline without relying on traditional software engineering teams. This democratization of development reduces reliance on expensive external consultants for internal data tasks and accelerates the delivery of analytical capabilities.
Heimstaden’s transition from manual property coordination to AI-assisted workflows demonstrates that the primary bottleneck in enterprise AI adoption is no longer the technology itself, but the readiness of underlying property data and the willingness to redesign end-to-end processes.
Related articles
