SHOPKEEPER

Track · Sell · Grow

A two-sided retail intelligence platform for micro-shops: fast offline sales terminal for staff, and an owner dashboard powered by operations research (reorder points, EOQ, purchase-window scoring) plus data-driven advice and cash-aware budgets — built for Orion Global Hackathon 2026.

E-commerce / Retail FinTech Machine Learning / AI INFORMS · Operations Research Offline PWA

Launch prototype GitHub

Problem & measurable impact

Small retailers lose margin to stockouts, overstock, and cash tied in the wrong SKUs. ShopKeeper connects every sale to inventory and profit, then recommends when and how much to reorder under cash constraints.

How judges can demo (3 min)

Launch prototype → first visit: create shop + owner PIN.
On the terminal: Reset Demo, record sales (−), pick active seller.
Open Owner Dashboard → health score, 7-day chart, OR reorder plan, purchase windows, budget.
Run Scenario Simulator with a price change to see elasticity-aware profit.

Innovation & OR (INFORMS)

Reorder point: R = μ_dL + zσ_d√L
Simplified EOQ for order quantity under holding cost
Purchase-window engine (Python + JS): discount vs 30-day mean, trend slope
Cash-constrained weekly budget allocation

Aligned with judging criteria

Innovation (25%) — OR + FinTech in one offline-first micro-retail stack
Technical excellence (25%) — PWA, modular analytics, Python backend roadmap
Real-world impact (20%) — Targets tuck shops & informal retail cash flow
UX (15%) — Clerk-fast terminal; owner depth without clutter
Demo (15%) — This page + live metrics from your own sales

Architecture

Browser PWA (localStorage) → or-engine.js + analytics.js → owner dashboard. Cloud path: PostgreSQL schema in backend/schema.sql, supplier intelligence via purchasing_engine.py.

[Sales Terminal] ──transactions──► [Analytics + OR Engine]
        │                                      │
        │                                      ├── Health / Forecast / Budget
        └── Staff PIN roles                    ├── Reorder & purchase windows
                                               └── Scenario (elasticity)
Future: API sync ──► PostgreSQL + supplier price history + embeddings