AI agents handle tenant requests

Property management is increasingly using AI agents to manage tenant requests, such as repairs and follow-ups, in order to save time and improve service. These AI tools are designed to handle specific tasks, tailored to the procedures of each property management agency.

Handling requests with AI

AI agents start by reading tenant messages, determining their nature and urgency, and directing them to the appropriate person. For example, an active leak is prioritized over a question about a rent receipt. If the agent answers common questions, it must identify itself as an AI, as required by the AI Act since August 2, 2026.

Managing repairs with AI

A second AI agent manages repairs by determining who is responsible for the costs—tenant, landlord, or insurer—and prepares requests for approval when necessary. This helps streamline the process and ensures that repair requests are handled efficiently and according to the correct procedures.

Tracking transfers and departures

A third agent handles property transfers by matching them to the correct lease, identifying any delays, and preparing follow-ups for each stage. A fourth agent manages tenant departures by comparing entry and exit inventories, calculating justified withholdings, and preparing restitution letters within legal deadlines. These tasks are often time-consuming and require careful attention to detail.

Preparing management reports

A fifth AI agent writes notes that accompany management reports, summarizing the quarter's facts and upcoming deadlines, such as rent reviews or energy efficiency thresholds. Each agent resolves specific questions and prepares documents, which the manager then proofreads, corrects, and decides to send. The calculations remain with the management software.

Improving landlord perception

The time saved by these AI agents is noticeable to landlords, who often do not see the work of their managers but notice silences and delays. AI agents change this perception by ensuring timely responses to tenant requests, providing repair approvals with quotes and reasons, and commenting on management reports rather than presenting them as a series of lines.

Notifying about important deadlines

Landlords are also notified well in advance about important deadlines, such as the energy efficiency threshold that will exclude class F properties from January 1, 2028. This helps them prepare for upcoming changes and avoid potential penalties or issues with their properties.

Challenges in AI implementation

Building AI agents requires specialized skills and integration with existing tools, including management software and messaging systems. These agents must be connected to the management software, whose technical access is often reserved for the publisher's partners, and to the messaging system. They must also translate procedures into rules and be tested on real cases before sending any letters.

Data and privacy concerns

The data involved, such as leases and tenant communications, are personal data, requiring tools based on professional offers with contractual confidentiality commitments. Agencies must know where this data is processed. These skills are rarely found internally, similar to those of an accountant or a web developer.

Gradual implementation of AI

The correct method is gradual, starting with one or two agents for the most time-consuming tasks, measuring the hours saved and errors avoided before expanding to other tasks. Property management does not lack landlords to serve; it lacks time to serve them well. Agencies that build their agents now will return this time to the relationship, making the difference in the industry.