Project History
Solar Max · Ridge Max Roofing · Velocity Air Conditioning. September 2021 – April 2026. Here's what I actually built.
Revenue Generated
$5M+
From one automation connecting two companies
Monthly Automation
19,000+
Automated tasks running at peak
Dialer Revenue
$1M+/yr
From one $150/month seat license
Tenure
4.5 years
Sept 2021 – April 2026
CRM & Operations Infrastructure
CRM Build from Zero
Designed and built the entire CRM infrastructure from scratch across three companies. Custom pipelines, automation rules, lead routing, and reporting. No template — built to match how the business actually operated.
Foundation for everything elseCross-Company Data Integration
Built an automation that connected the data of two separate companies — Solar Max and Velocity Air Conditioning — creating a $5M revenue pipeline that didn't exist before. One workflow. Real money.
$5M revenue pipelineKPI Dashboards
Built Google Sheets dashboards pulling live data from the CRM and operational systems. Leadership could see revenue, pipeline health, job completion rates, and team performance in real time.
Real-time visibilityAutomation & AI Systems
19,000+ Monthly Automations
Built and maintained the full automation stack — Zapier, Google Apps Script, and eventually MCP-based integrations. At peak: over 19,000 automated tasks running every month across all three companies.
19,000+ tasks/monthAI Call Quality Assurance
Deployed an AI system that listened to sales and service calls, scored them against defined criteria, and flagged issues for review. Replaced manual call auditing entirely. Consistent, scalable, and faster than any human review process.
Replaced manual auditingAI Photo Inspection — Job Completion Compliance
Built for the HVAC division (Velocity Air Conditioning). Technicians submit job completion photos through ServiceTitan. Photos route to Google Drive, where an image hashing and AI review system checks for compliance before the crew leaves the job site.
HVAC compliance automationPredictive Dialer Campaign
Built and managed a predictive dialer campaign that generated over $1M per year in revenue — off a single $150/month seat license. The ROI on this one was hard to explain to people who hadn't seen it.
$1M+/yr from $150/moField Operations & Logistics
Vehicle Inspection System
Built a weekly vehicle inspection process that starts every Monday. Technicians complete digital inspection forms in the field. Results escalate through the week — reminders, manager alerts, and compliance tracking — until resolved.
Weekly compliance processHVAC Truck Inventory System
Built a weekly inventory tracking system for HVAC service trucks. Technicians submit inventory counts via form. System generates a pull list, flags low stock, and sends restocking requests with escalating warnings if not addressed.
Automated restocking workflowSamsara Vehicle Telematics Integration
Samsara charges $30+ per vehicle per month for data exports. With 50+ vehicles on the fleet, that was a significant cost. Built a custom workaround that pulled the same telematics data into the reporting stack at no additional cost.
50+ vehicles, $0 extra costTechnician Scheduling & Dispatch Automation
Built automated scheduling workflows that reduced manual dispatch time and improved technician utilization. Jobs routed based on skill set, location, and availability — with customer notifications triggered automatically at each stage.
Reduced dispatch timeDigital Infrastructure & Marketing
Digital Marketing Infrastructure
Contributed to the digital marketing strategy and was part of the process for building tracking infrastructure, campaign attribution, and lead source reporting. Focused on making sure the data behind the marketing was clean and actionable.
Tracking & attributionMonday.com Implementation
Core contributor to the Monday.com rollout across the organization. Built approximately 60% of the boards, automations, and integrations. Connected Monday.com to the CRM and operational data sources so project status reflected reality.
~60% of buildIT & Infrastructure Support
I did all of it — and when something needed to be connected, moved, or wired up, that was me too. Workstations, phones, network connections — whatever it took to keep operations running. I had someone handling the internal hardware on the PCs, but everything else that needed to be plugged in, set up, or figured out? That landed on my desk.
Full operational coverageActive Consulting Clients
AMG AI Solutions has two active consulting clients. These are side projects — not the main focus — but they're real, ongoing, and worth mentioning.
IGBT Parts Distributor
My cousin's business — the one that started this whole thing. Selling industrial components globally. Built integrations connecting QuickBooks, Shopify, and warehouse inventory in Google Sheets. Started on weekends in 2024. Still running.
Warehouse Inventory Client
Connected through an AI mixer event. Building inventory automation and reporting workflows for a warehouse operation. Early stage, but the problems are real and the solutions are working.
How my thinking evolved
Three generations of integration thinking, lived in sequence. Each one made the previous one look slow.
2021–2023
Zapier
No-code automationDrag and drop. 19,000 tasks a month. Powerful for its time. But every integration had a cost, a rate limit, and a layer of abstraction between me and the actual data.
2024
Google Apps Script
Direct API accessCut out the middleware. Talked directly to APIs. Faster, cheaper, more powerful. Realized I could build things Zapier couldn't touch — and do it for free.
2025–Now
MCP + AI
Native AI integrationThe integration layer is disappearing. MCP connectors let the LLM understand the data structure itself and query it natively. Point it at a source, describe what you want, it figures out the rest.