Heimstaden has begun a broad push to embed AI in everyday operations as it seeks to modernise how it markets, manages and assesses its residential portfolio. Central to that effort is AI property management: the company accelerated its programme in June 2025, selected OpenAI after evaluating multiple platforms, and says the objective is to empower staff, redesign core processes and keep humans responsible for final decisions.
The rollout touches several stages of letting. Heimstaden uses ChatGPT Sites alongside other OpenAI tools to inform pricing, generate styled images and draft listing copy. Content creation and market feedback have been automated to make listings more responsive to demand, while teams continue to review and approve outputs against internal quality standards. The company reports particularly striking results from virtually staged imagery in the Netherlands, where production costs fell from about €50 per image to under €0.10 — a reduction of more than 99%. Heimstaden estimates that saving alone could translate to roughly €130,000 in annual savings in that market.
Heimstaden organises its digital approach around three pillars: equipping staff with appropriate tools and training, coordinating initiatives through a central transformation office supported by AI champions, and preparing property management systems and data for AI use. The central transformation office also maintains a library of approved use cases so teams have practical examples to follow as they adopt new tools.
The company is enabling domain experts, not just engineers, to build analytical capabilities. Using Codex, an environmental sustainability analyst located, cleaned and standardised scattered public data to produce a portfolio climate-risk assessment. That work produced a dataset covering 20,112 assets and more than six million rows of structured data. Heimstaden says the use of Codex reduced delivery effort dramatically: what would have required five people working for a year was completed by two people in two months. The project avoided external consultancy, produced a repeatable process and created material Heimstaden can apply to reporting and investor decision support.
Heimstaden emphasises responsible adoption. Employees remain responsible for reviewing AI outputs and approving final decisions. When ChatGPT Enterprise launched in September 2025, roughly half the workforce adopted the tool within two to three months of rollout, the company reports. Adoption has extended beyond corporate teams: caretakers and maintenance staff are using AI in operational tasks. In Norway, for example, a caretaker used ChatGPT to assess a neglected flat and turn informal observations into a prioritised refurbishment plan for review.
Beyond standalone tools, Heimstaden is integrating models into workflows that span its business systems. Its Procurement Intelligence Platform, currently being rolled out, combines workflow automation with OpenAI models accessed via APIs. Designed to work with Outlook, SharePoint and Teams, the platform assists with contract analysis, negotiation preparation and contract lifecycle tasks while leaving commercial decisions and approvals to employees. In parallel, Heimstaden is co-developing Amplio Estate, a modern property management system intended to support AI-driven workflows through an MCP integration so employees can use models such as ChatGPT to retrieve property information and trigger actions in core systems under human oversight.
The company cites several concrete results from the programme: the greater than 99% reduction in image production costs in the Netherlands; the Codex-enabled assessment of 20,112 assets; and about 50% employee adoption of ChatGPT Enterprise within two to three months of its launch. Heimstaden also highlights three leadership lessons from its rollout: prioritise people because adoption and imagination are primary constraints; identify local champions to drive change; and ensure processes are sound, since AI scales good processes rather than fixing broken ones.
Heimstaden frames these changes as steps toward buildings supported by increasingly intelligent systems while keeping people as decision makers. By reducing operating costs, speeding routine decisions and building repeatable analytical capabilities internally, the company demonstrates a path for a large residential landlord to deploy AI at scale without outsourcing core analytical work. As Heimstaden moves from point tools to models that act across systems, its emphasis on staff training, governance and clear approval responsibilities is intended to keep human oversight at the centre of AI property management.
Source: Read the original source
Leave a Reply