Overview

We built a chat-based AI assistant to reduce repetitive and manual email-based work of managers in Condo/STRATA buildings. We trained an LLM (large language model) on thousands on internal documents and local by-laws to instantly address most resident questions and reduce communication workload by 50%.

Challenge

The client was a Property Management company with a portfolio of 15,000+ urban multifamily units across British Columbia, Canada. Condo/HOA managers are getting overwhelmed by the amount of resident communications, resulting in burnout and low quality of service. Most of this email traffic are simple, repetitive things like rule clarification, document requests etc.

Approach & Solution

By conducting in-depth discovery sessions with client’s team and analyzing thousands of resident emails, we mapped their communication traffic into higher order categories.

We built a chat-based LLM and trained it on thousands of internal documents and by-laws to accurately address a vast majority of resident requests while also giving managers real-time updates on urgency, status etc.

Impact

The AI assistant solution reduced email volumes by 50% and increased FTE capacity by 18%. The managers were able to resolve a majority of resident-council requests in a fraction of time compared and our solution gave the client a distinct competitive advantage by deploying industry-leading technology.