What a Brookshear Operations Engagement Looks Like

The owner of a 12-person HVAC company was spending 3 hours every Monday manually pulling job completion rates, revenue, technician utilization, and open invoices from three different systems — ServiceTitan, QuickBooks Online, and a Google Sheet — then assembling everything into a single spreadsheet report. That's 150+ hours a year lost to report-building instead of running the business.

Built a Python + Google Apps Script pipeline that connects directly to ServiceTitan (their field service software via REST API), QuickBooks Online, and Google Sheets. The script pulls job completion rates, revenue by technician, open invoices, and utilization metrics, then feeds them into a single Looker Studio dashboard configured with the owner's KPI layout. A cron schedule triggers the full refresh automatically every Monday at 6am — the dashboard is live and waiting before the owner sits down.

New client inquiries came through the agency's website form and landed in someone's email inbox. From there: manually copy the lead into HubSpot, manually assign to the right account manager, manually send a welcome email, and hope nobody dropped the ball. Four manual steps, 20–45 minutes per lead, and leads were routinely falling through the cracks when the team was busy.

Built a Zapier multi-step workflow that handles the entire intake sequence the moment a form is submitted: HubSpot contact is created automatically with all fields populated, conditional routing logic reads the service type selected on the form and assigns the lead to the correct account manager, a personalized welcome email goes out under the AM's name within seconds, a Slack notification fires to the assigned AM with lead details, and an Asana task is created for the discovery call follow-up with a deadline pre-set to 48 hours. Zero human intervention required.

A 22-person commercial cleaning operation receives 40–60 vendor invoices every month as email attachments. One admin was spending approximately 8 hours per month manually opening each PDF, reading the vendor name, invoice number, line items, amounts, and due dates, then re-keying all of it into QuickBooks Online. Beyond the time cost, the manual entry averaged 3–4 data entry errors per month — resulting in mispaid invoices, duplicate payments, and month-end reconciliation headaches.

Built an AI-powered document processing pipeline using Anthropic's Claude API to read and extract structured data from each PDF invoice — vendor name, invoice number, individual line items, amounts, and due dates. A Python script takes the extracted data and validates it against the company's vendor master list, flagging anything that doesn't match a known vendor or falls outside expected amount ranges. Clean invoices are automatically pushed to QuickBooks Online via API as draft bills ready for one-click approval. Exceptions get routed to a review queue for the admin to handle directly — the system handles the routine 90%, the human handles the edge cases.

A 3-location HVAC services owner had data scattered across five systems — QuickBooks Online (financials), ServiceTitan (jobs & CRM), Stripe (customer payments), a bank account (cash position), and Google Workspace (emails & documents). To answer a basic question like "which technician is most profitable this month?" or "what's my cash position vs. AR?" required pulling reports from each system manually, then cross-referencing in a spreadsheet. The owner gave up and was running the business on gut feel. By the time monthly reports landed, they were 6 weeks out of date and missing the cross-system context that actually mattered.

Built an automated, real-time operations dashboard that aggregates data from all five source systems. Webhooks from ServiceTitan and Stripe push job and payment events in real time. QuickBooks Online and the bank account sync every 15 minutes via API. A Make.com scenario normalizes the data into a common schema, then Google Apps Script transforms it into pre-calculated KPIs (revenue, gross margin, AR aging, technician scorecards, top customers, cash runway). The dashboard renders the live data in a single owner-facing view that refreshes automatically. Scheduled reports (weekly summary, monthly close, margin alerts, AR reminders) fire automatically and email the owner without anyone touching a spreadsheet.

Beyond just visibility, the dashboard runs 4 automated workflows in the background: (1) Weekly operations summary emails delivered every Friday at 4pm to the owner's inbox — pre-formatted P&L, top customer breakdown, technician scorecard. (2) Monthly close automation generates a complete close package by the 15th of each month. (3) Margin alert monitor watches all technician and job-type margins continuously — fires an email alert if any drops more than 5 points below the team average. (4) Automated AR reminder sequence sends customer payment reminders at 30, 45, and 60 days overdue without anyone touching the inbox.