Systems in production.
Real clients. Real results. Every build here is running right now – not a demo, not a prototype.
250 qualified roofing leads, every Monday.
The problem: Top Dawg needed a steady stream of commercial roofing prospects across southern Alberta – property managers, facility managers, building owners. Manual research wasn't scaling.
What was built: A 5-source parallel fetch pipeline – Google Maps Places API, Apollo.io (companies + contacts), Calgary/Lethbridge/Red Deer building permits, BOMA member directory, and Alberta Corporate Registry. All sources run simultaneously, merge into one schema, deduplicate against the master sheet, enrich missing emails through a 3-step waterfall (Apollo → Hunter → Pattern + Million Verifier), pass through a quota gate, and push to Google Sheets.
The result: 2,000-lead baseline seeded in a single run. 250 new verified leads every Monday with a Slack summary – source breakdown, dedup count, enrichment stats. The sales team works from the sheet; the pipeline runs without them touching it.
Website, leads, and SEO — all running on autopilot.
The problem: Inline Mechanical had no web presence, no consistent way to source new commercial leads, and no visibility into how they were performing in local search. Three separate problems that each needed a system, not a one-off fix.
Website: A 7-page marketing site (HTML/CSS/JS + GSAP) deployed to Netlify with a contact form, service pages, and conversion-focused copy. Live in days, no CMS overhead.
Lead sourcing: Apollo-based pipeline targeting GCs, homebuilders, strata managers, and commercial property contacts across BC. Enriches via Apollo → Hunter → MillionVerifier waterfall and pushes verified leads directly to Google Sheets.
SEO agent: Pulls 28-day data from Google Search Console every Monday — clicks, impressions, keyword positions, and page performance. Flags position-5-to-15 quick wins, audits top pages for thin content and missing meta, and outputs a prioritised 3–5 item action list to Google Sheets. No monthly reports that sit unread — concrete tasks with the page or keyword named.
3+ hours of admin work handled before 9 AM.
The problem: Inbox triage, CRM logging, pipeline pulls, outreach drafts – all happening manually every morning. High-judgment but low-value admin that ate the first 3 hours of every day.
What was built: A Claude-powered EA that runs a morning routine: triages Gmail inbox (labels + summaries), pulls Instantly campaign replies, logs call summaries to HubSpot, drafts outreach emails, fetches calendar, and delivers a structured morning briefing to Slack before 9 AM.
The result: Day starts with a Slack briefing instead of inbox firefighting. Outreach drafts are queued and ready for review. CRM is updated from call notes automatically. 3+ hours of admin reclaimed daily.
JY Solutions' own LinkedIn content, automated.
The problem: Creating LinkedIn content consistently takes time that doesn't exist. Most content pipelines produce generic posts that don't sound like the person they're supposed to represent.
What was built: An 8-step pipeline with an embedded JY voice profile – construction vocabulary, direct tone, no emojis, specific numbers over vague claims. Transcript goes in via Fathom API or file. Posts come out mapped to content pillars, with carousel outlines and seed comments, scheduled on a Tue/Wed/Thu cadence.
The result: One 45-minute conversation → 6-week content calendar. Posts pass a quality audit before export. Blueprint filter catches posts that need restructuring. Seed comments are generated automatically for the first-hour engagement window.
What's the process
you'd automate first?
Free 30-minute call. We scope it, price it honestly, and figure out if it's worth building.