Iggy Artemenko wanted to go from 0% to 40% of new signups prepaying. The AI gave him an 8% discount and a price lock, and nothing else.
A plumbing-company owner split grease-trap sales between an AI researcher and an AI follow-up desk
Turn an approved list of restaurant prospects into a source-backed research sheet and draft-only follow-up queue.
What they did
An owner on X described using ChatGPT and Claude to work out a process, then Grok Bot for recurring tasks. For one plumbing company’s restaurant grease-trap pumping service, one agent researched commercial kitchens and another handled replies and follow-up, with a shared handoff area.
How it went
The owner reported about $10,000 in recurring restaurant revenue and about $100,000 in contracts from a separate property-manager team. The post does not state a clean measurement period or provide a contract ledger. Those are the owner’s claims, not verified returns. The transferable idea is separating research, conversation status and exceptions so a lead does not disappear between steps.
Worth knowing
The source allows some standard replies to go out automatically. Our adaptation is draft-only: an owner approves every message. Verify contact details, permission, opt-outs and applicable outreach rules. Do not invent local pumping requirements or imply an inspection found a violation. This is a firsthand social-media account or discussion, not an independently verified business result. The reusable prompt is our adaptation, not a quote of the poster’s exact setup. Wisconsin angle: A Wisconsin plumbing business can begin with existing commercial contacts and approved service areas. Use its actual service capabilities and availability, not a generic claim that every restaurant needs the same schedule.
Try it yourself
Want help getting this working? Rise Systems in Oconomowoc sets these up.
Want this running in your business without the DIY? Rise Systems in Oconomowoc sets these up.
Free 15-minute AI check-upSource: X (Twitter) · October 13, 2026
Create a small batch of useful local posts from approved job facts, with every claim checked before publishing.
Find useful customer questions for one service page, then draft only answers supported by your actual policies.
A counter-story. His site said "no AI." The code lives in a block he can't follow, and it has problems on mobile.
Get one idea like this every morning.
Follow on Facebook