A $3M+ business running on gut feel and manual work
A commercial painting contractor in the Southeast — 140+ active jobs, mix of residential and commercial work, growing fast. JobTread and QuickBooks both active and paid for. But almost nothing automated.
Estimates built from scratch every time. Vendor invoices re-keyed by hand. AIA billing assembled manually each cycle. Job profitability only visible after a job closed — when there was nothing left to do about it. The owner was asking the office for updates that should have been on a dashboard.
"They were paying for two good platforms and getting maybe 20% of the value. The data was there — it just wasn't connected to anything useful."
The foundation before any tools could be built: a full catalog restructure. Every cost code, product, and service rebuilt so data flows cleanly from Dext to JobTread to QuickBooks without manual correction at each step. That work made everything else possible.
AI Estimating — scope description to JobTread budget in under two minutes
Every estimate was being rebuilt from scratch. Estimators had to manually search the cost group catalog, identify applicable line items, guess quantities, and create each cost group one by one. Unknown items got skipped. The catalog never improved.
I built an AI estimating tool connected directly to their active JobTread cost catalogs. Describe the scope by voice or text — the tool matches every task to the correct catalog entry, estimates a realistic quantity, and pushes the full budget into JobTread as a draft with one click. Anything it cannot match gets emailed to the manager as a catalog improvement request.
Outcome: Estimate creation dropped from 20-40 minutes to under two minutes. Catalog-to-job matching is exact — IDs, not guesses — so cost reports are clean. Every flagged item becomes a documented catalog improvement, meaning the tool gets more accurate the more it is used.
Financial Overview Dashboard — month-end review in 30 seconds
Reviewing monthly performance meant pulling a manual QuickBooks report and interpreting raw numbers with no benchmarks and no color signals. Nobody agreed on what "on target" looked like.
I built a live dashboard pulling QuickBooks P&L data in parallel, with benchmark thresholds baked in — COGS%, GP%, and Net Profit% displayed as color-coded pills. Green when hitting targets, red when not. No login to QBO required.
Outcome: Month-end review that previously required navigating QBO and manually interpreting figures now takes under 30 seconds. Every stakeholder works from the same benchmark thresholds.
Job Scorecard — cost burn, GP%, invoicing, and AR for any job in one view
To understand how a specific job was performing, someone had to navigate multiple screens in JobTread and piece the picture together manually. There was no single view.
The job scorecard pulls live data from JobTread for any selected job — cost burn vs. budget, projected GP%, invoicing status, and AR outstanding — all in one screen. Admin-only access.
Outcome: Job-level review that previously took several minutes of navigation now takes seconds. Underperforming jobs surface earlier — when there is still time to act.
Operational Dashboard — cost breakdown and revenue by job type, always current
There was no way to see cost breakdown by type (labor vs. materials), revenue by job type, or operational trends across the full portfolio without building a custom report from scratch every time.
The operational dashboard aggregates live JobTread data into a single view — cost breakdown, labor vs. materials split, revenue distribution by job type. Admin-only. No exports required.
Outcome: Operational patterns that were previously invisible are now reviewable in one place. Cost-heavy job types and labor/materials imbalances are visible without custom reports, enabling faster decisions.
Sales Dashboard — live pipeline with rep filters, market filters, and urgency signals
The sales team had no shared view of the pipeline. Proposals sat in JobTread with no urgency signal — a job sent three months ago looked identical to one sent yesterday. Stale estimates slipped through because nobody was systematically watching them.
I built a live pipeline view with layered filters by rep and by market, with color-coded status badges on open proposals. Filters stack without replacing the base view. Every rep can see their own pipeline without asking a manager.
Outcome: Stale proposals are visible before they go cold. Rep accountability is managed from one screen. Fewer estimates die silently.
Follow-Up Tracker — a prioritized queue replacing memory-based follow-up
Sent proposals had no urgency signal. The sales lead had to manually dig through individual jobs to figure out who had not responded, when they last viewed the proposal, and what to do next. Follow-ups happened on gut feel — or did not happen at all.
The follow-up tracker queries JobTread for all jobs in "Estimate Sent" status with a qualifying proposal document. Each job surfaces with its issue date, last view timestamp, and email delivery status. Sorted oldest-first with color-coded urgency — green under 14 days, yellow 14-30, red 30+. Job status can be updated inline without leaving the tool.
Outcome: Follow-ups went from memory-based and inconsistent to systematic. Fewer proposals die silently. Conversion rate on sent estimates improved as a result.
Review Requests — automated trigger at job close, no memory required
After jobs closed, getting Google reviews depended entirely on someone remembering to send a request, find the customer contact, and write a message. Reviews were being left on the table at the exact moment customers were most satisfied.
The review request tool pulls closed jobs from JobTread, surfaces customer contact info, and sends a review request with one action. The trigger is the job close status — no memory, no manual step required.
Outcome: Review request rate went from inconsistent to systematic. More touchpoints with satisfied customers at exactly the right moment.
AIA G702/G703 Generator — compliant billing documents in minutes, not hours
Billing on commercial jobs requires AIA-compliant G702/G703 payment application forms. These were assembled manually every billing cycle — filling continuation sheets by hand, calculating stored materials and percent complete, then formatting a document the GC would accept. Slow, error-prone, and required specialized knowledge.
The AIA tool generates compliant payment applications by pulling job data from JobTread and walking through the billing period details. It outputs a properly formatted document ready to submit to the general contractor.
Outcome: Billing documents that used to take 30-60 minutes now take minutes. Fewer errors, faster cash flow on commercial work, and no specialized AIA formatting knowledge required.
Vendor Invoice Processor — Sherwin-Williams, Home Depot, and EquipmentShare handled automatically
Invoices from vendors arrived as PDFs with no connection to job costing or accounting. Someone had to manually read each one, identify the job, and key the data into both JobTread and QuickBooks by hand. Different formats from different vendors made it inconsistent at scale.
The vendor invoice processor accepts any supported vendor PDF and routes it through an AI extraction pipeline — identifies the vendor, pulls line items, and matches the invoice to the right job and budget line. The operator reviews and confirms, then with one action: the vendor bill is created in JobTread, the expense is pushed to QuickBooks under the correct account and class, and the payment is recorded. Sherwin-Williams, Home Depot, and EquipmentShare each have dedicated cost code mappings built in. The architecture supports onboarding additional vendors without rebuilding the pipeline.
Outcome: A recurring multi-platform manual task is now a single lightweight checkpoint per invoice. Job costing stays current in JobTread automatically. Three vendors supported from day one, with the foundation in place to add more.
From manual everything to a full ops platform in 8 weeks
"The goal was not to replace what they were doing. It was to make it actually work — and give the owner visibility he could act on, not just data he had to go looking for."
What is your team still doing manually?
If you are on JobTread or QuickBooks and your team is still doing by hand what should happen automatically, let's talk. One 20-minute call — I'll tell you exactly where the gaps are and what is possible, and what a retainer would look like for a business your size.
Or email directly: andres@tradeopslab.com