AI Agents, PM Policy Tracking & the Pod Hybrid Model with Wolfgang Croskey

Mar 26, 2026

Wolfgang Croskey, co-founder of Crane and a property-management technologist, explains building on‑prem AI agents like Bruce and Sophia. He discusses running agents on VPS, integrating SMS/Slack/voice, and a local policy‑tracking service that watches city and county meetings. Conversation also covers CraneHQ tools, automating outreach, and tradeoffs in org design for property managers.

Ask episode

AI Snips

Chapters

Transcript

Episode notes

ANECDOTE

Homegrown AI Agents That Run On Your Server

Wolfgang built two in-house AI agents, Sophia and Bruce, running on a VPS to triage emails, pull AppFolio data, and delegate tasks via Slack.
Bruce triages LeadSimple emails, runs cron jobs with AppFolio read-only access, and nudges teammates when tasks stall.

ANECDOTE

AI Pulled Competitor Listings And Built Owner Leads

Wolfgang used OpenClaw on a Mac Mini to monitor a competitor's listings and auto-scrape owner contact info after a listing sat 28 days.
The workflow produced a spreadsheet for targeted outbound mail and owner outreach.

ADVICE

Limit AI Access To Read Only And Internal Channels

Keep integrations limited to APIs and read-only connections to protect sensitive data.
Wolfgang moved AppFolio access to read-only after a prompt-injection scare and restricted Bruce to internal-facing channels.

Get the Snipd Podcast app to discover more snips from this episode

Get the app

🔗 Register for Crane's Access Granted launch (April 2)

This week on Peter Lohmann's Podcast:

What if your property management company had its own AI agents?... Ones you built yourself, running on your own server? Crane co-founder Wolfgang Croskey built two custom AI agents (Bruce and Sophia) that triage emails, delegate tasks via Slack, and pull AppFolio data, all on his own server.

Chapters:

(00:00:00) - Intro
(00:03:09) - Meet Sophia and Bruce
(00:04:23) - OpenClaw workflow wins
(00:06:03) - Building agents on VPS
(00:12:36) - Texting and voice assistants
(00:18:03) - Crane HQ origin story
(00:24:32) - Data providers and costs
(00:29:00) - AI policy email alerts
(00:31:33) - Crane launch details
(00:36:53) - RL’s hybrid pod structure
(00:43:31) - Early results and tradeoffs
(00:50:09) - Phone and email setup
(00:53:03) - Tech gaps and next steps
(00:56:21) - Wrap up and outro

We get into the security scare that changed his approach, why messaging through your normal channels matters more than the AI itself, and what it takes to actually set this up.

We also reveal a new Crane member benefit: an AI-powered local policy tracker that monitors city council, school board, and county meetings across your service area — down to cities with 2,000+ people.

Then I break down something I haven't shared publicly: RL Property Management's shift from departmental to a pod hybrid model at 750 doors - the owner churn that drove it, how we structured one-person pods with centralized ops, and early results including two saves from churning clients this week.

If you're thinking about AI agents, org design, or giving owners the relationship they actually want, this one covers it.

🔗 Wolfgang Croskey on LinkedIn

Resources for Property Managers & Real Estate Entrepreneurs

• Crane – Private PM Owner Community → Join a private network of property management owners and operators

• Free TWICE Weekly Newsletter → Property management insights, strategies, and industry updates direct to your inbox

• RL Property Management → Learn more about Peter’s company and services in Columbus, Ohio

Disclaimer: The content of this podcast is for informational purposes only and does not constitute professional advice. I may have consulting agreements with, or financial interests in, companies mentioned in this podcast. Additionally, some of the links included may be affiliate links, meaning I may earn a commission if you purchase through these links. Always perform your own due diligence before making any financial or business decisions.