Real engagements.
Real outcomes.

RECO sells recovery equipment (massage guns, cupping, compression sleeves, saunas, cold plunges) to gyms, schools, teams and retailers, and direct to consumers through Shopify. Two founders, no engineering staff, so building a system themselves wasn't an option. Follow-ups from meetings lived in notes and got lost.

I built one dashboard around their real data, then added agents that do the watching for them. Wholesale accounts, leads, inventory and events load from the sheet, plus a daily Shopify sync for inventory and DTC orders. Every account and lead can be edited, snoozed or flagged, with a full audit trail.

A single Priority Action Queue collects what needs attention today. Approving a drafted email reply sends it as a real threaded Gmail reply; approving a scheduling item creates the real calendar event. Agents flag inbox threads left unanswered for 3+ days, turn meeting notes into action items, find outreach and blogger opportunities, and draft posts for RECO's Shopify blog. Every finding has to cite a real email thread or event, and a second model checks tone and accuracy before anything reaches the queue.

The dashboard also carries QuickBooks P&L, cash flow, balance sheet and aging on a Finance tab, and a two-way sync between the accounts list and Notion.

At the time, 25N ran five coworking locations across the Chicago and Dallas-Fort Worth metro areas. Occupancy lived in a separate space-management platform that nobody looked at next to the financials, and leadership had no systematic read on how they compared to regional competitors.

I started with the question leadership could act on fastest: where do we stand against competitors? A competitive analysis covering five direct rivals, local SEO, Google Business Profile health and review velocity was delivered 9 hours after kickoff.

Then I built the finance side, one piece at a time, with the finance team reviewing every round. A GL dashboard put every location's P&L in one place, with revenue and NOI against budget, month-over-month variance flags, a six-month trend, run-rate projections for partial months, and a plain-English AI summary for each location whenever new data landed. A GL check applied per-account rules to flag entries that looked wrong before the close.

The accounting system already emailed daily GL exports, so a watcher picked them up and loaded them automatically. Month-end financial packets (income statement against budget, balance sheet, cash flow, AR and AP aging, occupancy) generated per location as PDF and Excel. Occupancy was pulled daily for every location and locked automatically at month end so close numbers didn't drift.