Projects / Real estate intelligence
AI for real estate intelligence
Automated insights for modern property management.
Keywords
LLM / AI
Natural language processing
Automated classification
Predictive insights
Real-time analytics
Property data intelligence
Introduction
Our friends at Homepal are transforming how real estate organizations work with data - not by adding yet another analytics layer, but by delivering ready-to-use metrics, KPIs and dashboards, built on real operational needs.
Many property teams today are stuck cobbling together monthly reports from their ERPs, juggling Excel sheets and trying to align numbers between departments. Homepal simplifies all of that by integrating directly with industry systems, offering a shared interface where finance, operations and leasing can work from the same data. Nightly updates, all automated and with no manual work required.
What sets Homepal apart is not just their technology, but the profound industry expertise shaping every dashboard, every metric, every integration. Whether it’s identifying long-standing maintenance bottlenecks, forecasting vacancy losses or supporting hiring decisions based on actual workload data, Homepal gives teams the visibility they need to act - not react.
When they approached us to explore how AI could take this even further, the challenge was clear: amplify what already works, by automating what still requires too much time and interpretation. That means moving from “what happened?” to “what should I do next?” by automatically flagging emerging issues, surfacing the most relevant KPIs and helping every user, from technician to CEO, get more from their data.
Challenge
Homepal’s platform collects vast amounts of property and maintenance data, but turning that data into actionable insights is time-consuming and complex. Property managers need a faster way to spot trends and prioritize issues without manually sifting through tons of data.
Tenant maintenance reports add another layer of complexity. The reports vary widely in detail and format, making it hard to detect urgent problems like water leaks or safety risks, and even harder to decide what needs attention first.
Goal
Simplify the lives of property managers by using AI to analyze data, surface the most relevant insights, explain the findings and deliver actionable recommendations where they matter most.
Solution
We approached the challenge by combining classical statistical methods for structured time series data with large language models (LLMs) for document and data analysis.
The statistical layer identified outliers, trends and future developments. Feeding those results into specifically configured LLMs allowed us to automatically highlight the most relevant information, explain it in plain language and present it in a digestible way - all with zero manual effort.
For tenant reports, the system automatically:
Identified the source of the issue (e.g. broken fridge, missing keys)
Detected signs of water leaks before they caused damage
Flagged safety concerns like poor lighting, broken windows, or suspicious activity
By applying LLMs across both structured and unstructured data, Homepal gained the ability to prioritize maintenance, detect urgent issues faster and see a complete picture of property performance and tenant concerns all within their existing platform.
Results & value add
The project proved that leveraging AI can meaningfully enhance real estate management by automating insight generation and aiding property managers in making faster and more informed decisions based on the data.
Key outcomes included:
Faster identification of issues from tenant reports, enabling quicker response times
Automatic classification of maintenance cases by type, urgency, and safety impact - a process previously done manually
Early detection of water leaks to prevent property damage
Data-driven insights from property data, reducing manual reporting and surfacing key trends
Seamless integration into Homepal’s existing platform, allowing insights to be acted on in real time