Why We Built a POS System for Local Retailers

When Shehroz and I started Solforge in 2021, we weren't thinking about AI or SaaS platforms. We were thinking about one thing: the small shop owner in Karachi who had a notebook and a calculator, losing track of inventory every single day.

The Problem We Saw

Walk into any local retailer in Pakistan and the story is the same. Sales are tracked on paper or in a basic Excel sheet. Inventory is a guess. End-of-day reconciliation takes an hour. Returns are a nightmare.

Enterprise POS systems like Square exist, but they're built for Western markets — the pricing, the payment integrations, the UI assumptions. Nothing fit.

"If you build something that solves a real problem for people you know, you've already won half the battle."

What We Built

We went with Flutter from day one. The reason was simple: one codebase for Android (the dominant platform in Pakistan) and iOS, with a native-feeling UI that could run on cheap tablets.

The stack:

Flutter for the tablet POS app
Supabase for the backend (real-time inventory sync, offline support)
Next.js for the web dashboard (owners check sales from their phone)
Stripe + local payment gateways for checkout

The offline-first problem

The hardest engineering challenge wasn't the UI. It was offline mode. A shop in a basement has bad connectivity. A power cut during a sale can't mean lost data.

We solved this with Supabase's real-time sync layered over a local SQLite database inside the Flutter app. Every sale is written locally first, then synced when connectivity returns. Conflicts are resolved server-side.

What We Got Wrong

We over-engineered the first version. Features like multi-branch inventory, supplier management, and employee roles — nobody asked for them. The shopkeepers we piloted with wanted three things: fast checkout, a daily sales report, and something their 16-year-old nephew could learn in an afternoon.

We cut 60% of the feature set before launch. The product got better immediately.

What It Taught Us About Building Software

Every project we take on now starts with the same question: what is the one thing this product needs to do really well? Not ten things. One.

That constraint is what turned a POS experiment into the foundation of Solforge.

If you're building something for an underserved market, we'd love to talk. The problems are harder, the competition is thinner, and the impact is real.