
How I built an Offline-First ERP that processed $30k in a market with spotty internet
Hi everyone, I wanted to share a journey of building a B2B product in a challenging environment.
I live in Egypt, where internet stability is a constant struggle for retail shops. Building a standard cloud-only SaaS for them was a recipe for failure. So, I spent the last few months engineering Manzoma ERP—a local-first system built with React 19, Electron, and Prisma/SQLite.
The Technical Challenge:
The biggest hurdle wasn't the UI; it was the Sync Engine. I had to ensure that if a shop owner makes 100 sales while offline, the system queues them perfectly and merges them with the cloud once the connection returns, without inventory conflicts across 3,000+ SKUs.
The Results:
- Processed over $30,000 (1.5M EGP) in transaction volume for my first few clients.
- Zero data integrity issues during offline-to-online transitions.
- 100% functional desktop app that feels like a web app.
Why I’m sharing this:
I’m heading into mandatory military service soon, which is forcing me to step away from the project. I’m proud of what I built as a solo dev using AI-assisted coding (Vibe Coding) to speed up the architecture.
I’d love to hear your thoughts on Local-first vs Cloud-only for B2B. Also, if anyone has experience exiting a "pre-revenue" but "high-volume" project, I'd love to chat!
Check out the landing page here: https://manzoma.online