← Cases
United StatesReal estate2025 – ongoing

AppFolio

Realm-X Messages, AI-suggested replies for property managers

AppFolio's Realm-X Messages saved property managers 11 hours a week on communications using Claude in Amazon Bedrock

WorkingVendor case studyVendor ↗

Problem

Property managers using AppFolio's software spend much of their week answering routine messages from residents, owners and vendors. Many of these messages need context such as resident history, open maintenance tickets and property policies.

Approach

AppFolio built Realm-X Messages, an embedded feature that suggests replies, on Claude in Amazon Bedrock. Claude 3 Haiku handles high-volume, quick replies to general inquiries. Claude 3.5 Sonnet handles complex messages that draw on resident history, maintenance tickets and property-specific policies.

Results

11 hourssaved weekly by property managers on communications, per AnthropicVendor case study[1]
26 secondsaverage reduction in message response time, per AnthropicVendor case study[1]

Takeaway

Splitting routine and complex messages between a fast model and a stronger one let the feature cover both. The company did not say how it measured the time savings.

Categories

Industry
Real estate
Business function
Customer service
AI technology
Generative AI
Evidence
Vendor case study

Sources

Source summaries

[1]Anthropic28 Sep 2025

Anthropic's customer story describes how AppFolio built the Realm-X Messages feature on Claude 3 Haiku and Claude 3.5 Sonnet in Amazon Bedrock. It reports 11 hours saved weekly by property managers and an average 26-second cut in message response time. The page carries no date; the Internet Archive first saved it on September 28, 2025.

Full details are in the original sources.

Added 4 Nov 2026 · Updated 4 Nov 2026
Open case page ↗