When I joined the team to build Store Sales, the scope was straightforward: a simple admin panel for managing products. Within months, it evolved into a full-blown enterprise SaaS platform supporting multiple retail branches, complex inventory workflows, and dozens of concurrent users.
The Challenge
Retail operations at scale are deceptively complex. A single branch needs to manage thousands of SKUs, process purchase orders from multiple suppliers, handle inter-branch product transfers, and generate real-time sales reports — all while enforcing strict role-based access controls.
Multiply that by multiple branches, each with different inventory levels and staffing, and you have a coordination problem that spreadsheets simply cannot solve.
Key Features I Built
Indent & Purchase Order Workflow — Branches submit indentation requests for stock. These flow through an approval chain before converting into purchase orders sent to suppliers. I designed the multi-step form UI with real-time validation to prevent duplicate requests and ensure data accuracy.
Material Receiving Reports (MRR) — When goods arrive at a branch, staff create receiving reports that are automatically matched against purchase orders. Discrepancies are flagged immediately, reducing inventory errors.
Role-Based Access Control — Different roles see different data. A cashier at a branch shouldn't access the admin's purchase order approvals. I implemented a permission matrix that controls navigation visibility, action buttons, and even API-level access checks.
Warehouse-to-Branch Transfers — Products move between warehouses and branches. The transfer module tracks quantities in transit, handles partial deliveries, and updates stock levels at both ends in real time.
Lessons Learned
- Start with the data model, not the UI. Understanding how purchase orders, indents, and receiving reports relate to each other saved weeks of refactoring.
- AG Grid is a lifesaver for enterprise tables. When you have tables with 20+ columns and thousands of rows, virtualized grids are non-negotiable.
- Optimistic updates improve perceived performance. Users don't want to wait for a server round-trip to see their changes reflected.
- Feature flags matter. Rolling out new modules to specific branches allowed us to test with real users before full deployment.
Store Sales is now actively used across multiple retail locations, processing real transactions every day. Building it taught me that enterprise software isn't about flashy UI — it's about reliability, data integrity, and workflows that match how people actually work.