I automated the NDA and CIM chase. Here's how it went.

A few weeks ago I posted about the geographic search and owner dependency problem. A lot of you shared really helpful perspectives. I want to share something else I've been working on that came directly out of the same frustration with the search process. Chasing brokers for NDAs and CIMs is one of the most repetitive parts of the search. You send a request, wait, follow up, wait again, chase down a form on some broker's website, then repeat it forty times. It's not hard work. It's just time-consuming work that adds up fast. So I built an AI agent to handle it. His name is Marcus. He runs on a Mac Mini in my office using an open-source framework called OpenClaw. Here's what he does: - Monitors my deal pipeline for new listings that match my buy box - Drafts personalized broker outreach emails after reading the listing - Sends NDA and CIM requests and follows up if no response in 48 hours - Navigates many of the NDA request forms on broker websites - When a CIM comes back, he flags it for my review with a quick summary I wouldn't trust him to evaluate a deal or make an offer. But the hours I used to spend on broker follow-up are mostly gone. I step in when it's time to review the CIM and take over from there. I recently hosted a webinar with about 30 other searchers walking through how I built Marcus and how anyone can set one up. I'm sharing the recording and the full build files here in case it's useful to the community. - Build files: joshua.thacker.com/openclaw - Recording: https://drive.google.com/file/d/1Yph0HIfYrtPM3zZLgUhQWu3aCLaHzLQC/view?usp=sharing A quick update on my search. Some of you will remember I was working a hard Sacramento radius. After two failed LOIs and the owner-dependency pattern that kept showing up, I started channeling the downtime between deals into what I love doing, which is building with AI. The pain of my own search is actually what led me to build Searcher OS, a CRM and operating system for the acquisition search process. That's where most of my energy is going now. I've even built a version of Marcus directly into the product that's in beta testing right now. Still looking at deals when they come across my desk, but the SaaS is the priority. I mention that partly for context and partly because I'm always trying to understand the biggest pain points in the search process so I can build better tools. If anyone would be open to a###-###-#### minute conversation about what feels broken or slow in your search, I'd genuinely love to hear it. Just trying to build useful things for this community. Happy to answer questions on Marcus, the AI setup, or anything else.