Introducing Shoppleet — The Offline-First Desktop Business Management Platform for East African Retailers

Last updated: May 2026 · By JB (Muke Johnbaptist) — sole developer and creator of Shoppleet, built on the Grit framework.

Most shop-management software in East Africa fails the same test: the moment WiFi drops, the till stops selling. That's an unacceptable defect when you're running a phone shop in Wandegeya at 6pm and the customer in front of you has cash in hand.

I built Shoppleet to fix exactly that. It's a native desktop business management platform — Windows, macOS, Linux — designed around one guiding rule: the till never waits for WiFi. Every sale lands in local SQLite first; the cloud syncs the moment a bar of signal returns. Cashiers never see a spinner.

After 18 months of building and 280+ shops running it across the region, today I'm publicly launching it.

---

TL;DR — what Shoppleet is

• A native desktop app (Windows / macOS / Linux) for East African retailers — not a packaged browser, a real binary.
• 9 modules: Dashboard, Sales, Purchases, Inventory, IMEI Tracking, Finance, Reports, Auditing, Administration.
• Offline-first via local SQLite + sync outbox — the till works without internet.
• Multi-tenant + multi-branch with strict per-tenant isolation.
• 175+ granular permissions across a 3-level catalog.
• Phone-shop-grade IMEI tracking — dual-SIM aware, lifecycle-locked, scan-to-resolve at the POS.
• Permanent audit trail + 3-12 month activity feed.
• Mobile companion app (iOS + Android) for owners on the go.
• 5 currencies: UGX, USD, KES, TZS, RWF.

---

Why I built it

Every shop owner I talked to in Kampala, Nairobi, and Dar had the same complaint about their existing software:

"It works when the internet works. When it doesn't, we go back to writing on paper. Then we lose a transaction. Then the stock count is wrong by Friday."

Cloud-only POS systems treat connectivity as a given. In our region, it isn't. Power flickers, fibre gets cut, MTN backbones reset, your daughter's phone hotspot dies — and your business stops earning.

The technical answer was obvious — make the app run entirely locally first, sync to the cloud opportunistically. The hard part was building it in a way that:

1. Handles IMEI tracking to phone-shop discipline (Uganda is a phone-retail powerhouse).
2. Survives a multi-cashier till where people sign in and out all day.
3. Produces audit trails that survive cross-examination — you can't delete a sale, only reverse it.
4. Generates print-ready reports owners actually want — not "data exports" that need an accountant to interpret.
5. Doesn't bankrupt the SME running on a UGX 800,000 laptop.

The result is Shoppleet — currently on v1.5.45, running in 280+ shops and tracking 1.4M+ IMEIs as of May 2026.

---

The "till never waits for WiFi" architecture

Every other claim about Shoppleet hangs off this one engineering decision: the database is local first, cloud second.

Cashier presses F4 (save sale)
        │
        ▼
Write to local SQLite (durable, instant)
        │
        ▼
Enqueue to local outbox table
        │
        ▼
Background worker drains outbox → api.shoppleet.com
        │
        ├─ Network up?    → Sync within ~2 seconds
        │
        └─ Network down?  → Stay queued, retry with exponential backoff
                            (next launch resumes from where it stopped)

What this buys you:

• Sales work offline indefinitely. Spend a day disconnected. Come back, the outbox drains.
• No "saving…" spinner at the till. F4 returns in <100ms because it's writing to a file on the same machine.
• Crash safety. If the laptop dies mid-sale, the next launch picks up cleanly — no lost transactions.
• Per-till idempotency keys mean a queued sale that already synced never double-posts.

The trade-off is sync conflicts — two tills selling the last unit at the same time. We resolve those with a server-authoritative stock allocator: whichever sale syncs first wins; the second gets a clear "out of stock — reverse?" prompt on its dashboard. No silent overselling.

---

The 9 modules at a glance

Every screen in Shoppleet lives in one of 9 numbered modules. The numbering matches what the cashier sees in the sidebar.

Each module gets its own permission tree, audit hooks, and printable reports — so a Cashier can ring sales but never view P&L, and a Stock Keeper can receive goods but never void a sale.

---

The IMEI tracking story (why phone shops chose us first)

Phone retail in East Africa runs on IMEIs. Without proper tracking you can't honour warranties, can't catch stolen returns, and can't reconcile when a unit "disappears" between branches. Most POS systems hand-wave this as "serial numbers" and leave shops to bodge it in Excel.

Shoppleet treats IMEI as a first-class citizen:

• Dual-SIM aware — captures IMEI1 + IMEI2 separately on every phone.
• Strict per-business uniqueness — the same IMEI can't be received twice in your tenant, even across branches.
• Lifecycle-locked status: IN_STOCK → SOLD → RETURNED / DAMAGED, never deleted.
• Luhn checksum validation on manual entry to catch typos at the point of input.
• Paste-from-Excel + CSV upload with duplicate detection (both within the file and against your existing stock).
• Auto-reject Excel-corrupted serials like 3.5679E+14 (anyone who's done bulk IMEI imports knows this pain).
• Scan-to-resolve at the POS — type or scan the IMEI, system finds the SKU automatically.
• Per-IMEI history page — every state change, who did it, when, why.
• Stolen / recalled device blacklist — a national registry hook so resold stock can't quietly slip back in.

That last point matters more than it sounds. As of May 2026, Shoppleet has tracked over 1.4 million IMEIs across customer shops — and the blacklist registry has already prevented dozens of stolen-phone resales.

---

The "no deletes, only reversals" doctrine

Every transactional record in Shoppleet — sales, purchases, expenses, transfers, stock adjustments — is append-only. You can't delete a sale to make a bad day look better. You can only reverse it with a required, operator-supplied reason. The reversal is itself a recorded row with is_reversal = true, the audit trail captures the operator, and the permanent audit log keeps the before/after JSON.

This is the single biggest reason owners trust Shoppleet to leave staff unsupervised:

"Before Shoppleet I had to count cash every evening to know if my cashier was clean. Now I read the audit trail in the morning. If someone reversed a sale, I see why." — Shop owner, Kampala (12-month customer)

The audit-trail screen has a financial-impact column so you can sort by "edits that cost money" and review only the suspicious ones. Compliance officers and SACCOs love this; cashiers respect it.

---

Built on Grit — the engine room

Shoppleet isn't built on Rails, Django, or Next.js. It's built on Grit — a full-stack Go meta-framework I designed specifically to generate every layer of a multi-tenant SaaS from one schema definition.

From a single Grit schema, the generator emits:

• Go API models, handlers, and routes (Gin + GORM)
• TypeScript types
• React Query hooks
• Desktop form scaffolds (Wails 2 + React 19)
• Mobile screens (Expo + React Native)
• Marketing copy stubs

That's how one developer (me) ships and maintains an end-to-end product — API, desktop client, mobile companion, and marketing site — without drowning in glue code. The Grit framework is my multiplier.

If you've ever wondered whether a Go meta-framework can power a real production SaaS, Shoppleet is the answer in production.

Stack snapshot

---

Multi-tenant, multi-branch — without the surprise bill

Shoppleet was designed multi-tenant from line one. That means:

• Strict tenant isolation at the API layer — no cross-tenant probing, ever.
• Per-tenant database scope on every query — even an admin in tenant A can't accidentally read tenant B.
• Multi-branch within a tenant up to the platform-set limit — branches share customers, suppliers, and product catalogue but track stock independently.
• Branch-to-branch transfers use a send→confirm flow — stock only moves on the receiving branch's confirmation, with a "missing in transit" report when math doesn't balance.

For a 4-branch retailer, that means the cashier at Branch 2 sees their till's stock, sales, and dashboard — never Branch 1's. The owner sees everyone's, consolidated.

---

Categories — the workflow adapts to the shop

Shoppleet ships category-aware UI. Pick your category at tenant creation and the right screens appear:

A phone shop doesn't see size×colour matrices; a fashion boutique doesn't see IMEI fields. The workflow stays sharp regardless of vertical.

---

What's NOT in v1 (be honest)

Shoppleet does a lot, but here's what it deliberately doesn't do yet:

• No online storefront / e-commerce — it's a back-office system, not a webshop builder.
• No online card-payment gateway — cash, mobile money, and bank work; online card processing for buyers isn't built. (For that, see DGateway.)
• No Jumia / Kilimall marketplace integration — coming, not shipped.
• No multi-currency on a single invoice — one currency per tenant. (Multi-currency across tenants works fine.)
• No QuickBooks / Xero export — on the roadmap, not yet shipped.
• No full payroll — there's a staff module, but proper payroll is separate.

If any of these block you, talk to Alex on +256 705 089 006 — we prioritise based on real customer requests.

---

The numbers (as of May 2026)

---

Who built it

Shoppleet was designed, developed, and built solo by me — JB (Muke Johnbaptist). Every line: backend, desktop client, mobile companion, marketing site, and the Grit framework underneath it. Alex Shanga supervises product direction and runs the commercial side — sales conversations, customer onboarding, support escalations.

---

How to try Shoppleet

• 14-day free trial — full functionality, no card required.
• Download at app.shoppleet.com (Windows, macOS, Linux).
• Cloud-hosted by default at api.shoppleet.com.
• Self-host option for enterprise customers — talk to JB / Alex.
• Mobile companion on the iOS App Store and Google Play (for business admins).

Onboarding is hands-on for the first 30 days — Alex or one of our partners walks every new customer through their first sale, first GRN, and first end-of-day report. No drop-and-pray.

---

What I learned building this

A few things I wish someone had told me before I started:

1. Offline-first is not "online + offline mode"

Most "offline" software is online-first with a degraded mode. Shoppleet is the opposite — online is the degraded mode of offline. The database is local; the cloud is a syncing peer. Every screen was designed to never know whether the network is up.

2. Audit trails sell themselves

Owners who try to negotiate on price stop negotiating the moment you show them the audit screen. "I can see who reversed which sale and why?" is a feature so valuable in our market that I now lead every demo with it.

3. Bulk import is 60% of customer satisfaction

A new shop with 4,000 phones in stock won't type each IMEI by hand. The paste-from-Excel + CSV upload + duplicate detection + auto-reject corrupted serials path is what gets them live in a day instead of a week. The bulk-import flow is the single most-debugged feature in Shoppleet. It earns its complexity.

4. One framework, many products

Grit lets me ship Shoppleet, DGateway, GritCMS, Vibekit, and a half-dozen client projects without burning out. The leverage of generating instead of writing is the difference between "founder who codes" and "team of 8 still missing features".

---

Related reading

• Desktop App Development in Uganda — HMK Estates Case Study — another Wails-based desktop app I shipped for the Uganda market.
• Desktop Frameworks 2026 — Electron vs Tauri vs Wails vs Grit — why I chose Wails for Shoppleet's desktop layer.
• Integrating DGateway into Next.js — the unified payments product that handles mobile money + cards.
• DGateway Framer plugin — drop-in payment components for Framer sites.
• Top Go Developers in Uganda 2026 — context on the Go-in-Africa scene that made Grit possible.

---

Want Shoppleet running in your shop?

Whether you sell 4 phones a day or run 8 supermarket branches, Shoppleet was built for you.

• 📞 Sales / demo / pricing: Alex on +256 705 089 006
• 💻 Download / try free: app.shoppleet.com
• 🛠 Technical / bugs / feature requests: JB on +256 762 063 160
• 🌐 Marketing site: shoppleet.com
• 📺 YouTube tutorials: @JBWEBDEVELOPER
• ⚙️ The framework underneath: gritframework.dev

---

Resources

• Marketing site: shoppleet.com
• App: app.shoppleet.com
• API: api.shoppleet.com
• Grit framework: gritframework.dev
• Wails (desktop framework): wails.io
• Asynq (Go job queue): github.com/hibiken/asynq