Learn more
How a Stockholm property developer freed 6,000 hours a year across development, production, and management
A Stockholm developer that builds and manages its own homes freed roughly 6,000 hours a year with AI built into every team. What we implemented, how each team learned to run it, and where the time went.
Titania is a Stockholm property developer that keeps the whole chain in-house. The company develops its own projects, builds them, and stays on as landlord, with hundreds of rental homes under management across the Stockholm region and more in production.
Leadership did not ask for automation for its own sake. They wanted their people spending less of the week on routine communication and paperwork, and more on the work that grows the portfolio. Finding sites, moving projects through planning, building, and letting homes. Our job was to build AI into how each team works, train people to run it well, and stay accountable for whether daily work actually changed.
We started by putting numbers on the routine load, team by team. Tenant service answered the same questions on rent, subletting, parking, and move-out by hand, and sorted every incoming fault report manually. During letting rounds, replies stretched to three days, in the exact weeks new tenants form their impression of their landlord.
Project managers collected contractor updates by phone and email, then retyped them into Friday status reports. Leasing rebuilt near-identical apartment listings from scratch for every letting round. Each team held knowledge the others needed, and there was no shared place where it lived.
A previous AI rollout had faded within weeks. People wanted the help, but generic tools with none of Titania's context could not carry real work. That set the bar for us. Whatever we built had to know the business.
We implemented the system first, then trained each team into it, and stayed on until it held without us.
- Mapped the routine load across tenant service, leasing, project management, and finance, and set the baseline numbers our work would be measured against
- Built one shared AI setup on Titania's own material. Lease terms, tenant guidelines, project documentation, and reporting templates, so every draft starts from the company's real context instead of a blank page
- Configured role-specific assistants for each team. Tenant replies drafted with the right policy attached, fault reports sorted and routed on arrival, apartment listings generated from project data, and contractor updates assembled into Monday status reports
- Ran a workshop series with every team on their own live cases, so training produced finished work rather than exercises, with follow-up coaching every second week until the habits held
- Connected knowledge across the company. What the site team knows now reaches tenant service answers, and what leasing hears from applicants reaches the development side
The point was never to replace judgment. It was to multiply it. Tenant service still decides every answer, but starts from a drafted reply with the right clause attached instead of a blank inbox. Leasing still shapes every listing, but from a generated first version. Project managers read status on Monday morning instead of writing it on Friday afternoon.
Tenants and applicants now get answers the same day instead of within three days, including during letting rounds. Nine in ten employees work with the setup daily. Counted across the company, the freed time comes to roughly 6,000 hours a year, and that time now goes into the developments that grow the portfolio. The setup keeps improving month by month on the retainer.
It became part of the team within weeks, not another system to manage. My people uses it to leverage their skillsets and are now much more productive and collaborative with their work. It is now the default first step in everything they do.
