Svenska Bostäder

Svenska Bostäder builds AI into real tenant services on Opper

svenskabostader.se

Industry
Public sector
apartments
28k+
Opper gives us a practical way to bring AI into real tenant-service environments. Instead of treating AI as something separate, we can connect it to the services and workflows we already use, supporting internal processes as well as image-based guidance in the tenant app.
Martin Larsson, IT Service Manager at Svenska Bostäder

About Svenska Bostäder

One of Stockholm's largest municipal housing companies.

Founded in 1944 and owned by the City of Stockholm, Svenska Bostäder manages thousands of apartments and serves tens of thousands of tenants across the city. As tenant expectations change, the company keeps looking for ways to make its digital services simpler and easier to use.

The Challenge

Bring AI into the services teams already use, not alongside them.

Svenska Bostäder didn't want AI as a separate experiment. The goal was to connect it to the systems and workflows teams already rely on, and to do that within the security, data protection and governance standards a City of Stockholm company has to meet.

The Solution

A developer platform for AI services, built on Opper.

The work started in 2024 with an internal AI interface hosted in Svenska Bostäder's own environment, where employees could work with different AI assistants for tasks like information retrieval, text support and internal guidance.

That same foundation now reaches tenants. In the tenant app, a resident reports a fault by sending a photo and a short description, and Opper processes that input, draws on different AI models depending on the task, and returns a structured response that supports the next step. Because these services have to meet the data protection and governance standards the company works to, the control plane gives the team oversight of how each one is built, measured and run.

  • Multimodal handling that reads a photo and a short description together to interpret a fault.
  • A choice of AI models through one gateway, so each task runs on the model that fits it.
  • Structured responses that drop into existing systems and shape the next step in the service journey.

Results

A practical foundation for AI-powered tenant services.

  • In production in the tenant app, where a resident reports a fault with a photo and a short description.
  • A growing set of AI tools, from internal assistants to tenant-facing guidance, built on one foundation.
  • EU data residency and GDPR-compliant handling, the standard a City of Stockholm company has to meet.
  • Room to build on, so new services can be tried and connected without starting from scratch each time.