Store Intelligence OS

Square sees payments. Toast sees orders. Clover sees terminals. Kolshee sees the entire business — and gives it away, because the data is more valuable than the rent.

Independent grocery stores run on a stack of disconnected tools: a POS for sales, a notebook for purchases, a paper schedule for staff, a folder for invoices, an Excel sheet for the P&L. None of them talk. None of them learn. The owner spends evenings reconciling what the day already knew.

Kolshee replaces all of this with one substrate. Every signal — sale, purchase, return, expiry, clock-in, utility bill, supplier invoice, shelf-weight reading — writes to the same event spine. The owner sees one screen. The mesh sees aggregated cohorts. Neither sees the other's noise.

01 · Product Intelligence

Tracks every product movement from purchase through sale, return, or waste. Velocity, margin per unit, preventability score on expiry events.

02 · Supplier Intelligence

Price history (12-month window), delivery reliability, invoice accuracy, quality score, network benchmark — this store's price vs network median.

03 · Financial Intelligence (Full P&L)

Revenue, COGS, gross profit, operating expenses, net profit, 30-day cash-flow projection. P&L statement auto-generated daily/weekly/monthly.

04 · HR & Workforce Intelligence

Employee records, attendance, shift management, labour efficiency (revenue per staff hour), payroll auto-calculated per period, performance scores.

05 · Operating Cost Intelligence

Rent, utilities, insurance, maintenance — every non-product expense tracked, anomaly-flagged, and benchmarked against similar stores in the mesh.

06 · Dead Space Intelligence

Dead shelf, overstock, underutilised storage, revenue per square foot. Recommendations: which shelf to rework, which SKU to replace.

Every store-level event is tenant-scoped by default. Cross-store intelligence operates exclusively on anonymised, aggregated cohorts: corridor-level price medians, demographic-cohort velocity benchmarks, supplier-performance percentiles. The store owner sees network benchmarks; the network never sees individual store financials. Employee PII is processed locally and never transmitted to the mesh layer.

• ▸Day 1: POS replaces existing terminal. Sales flow into the spine.
• ▸Week 1: Supplier invoices uploaded; purchase records reconcile against sales.
• ▸Week 2: HR onboarded; clock-in/out replaces paper schedule.
• ▸Month 1: First auto-generated P&L. First pooled-PO recommendation.
• ▸Month 3: Cross-store benchmarks active. First demand-gap recommendation accepted.
• ▸Month 6: Store owner stops reconciling at night. Spends evenings at home.