Full-stack engineer from Tunisia building operational products: POS systems, business dashboards, AI-assisted workflows, offline mobile apps, hardware integrations, and production infrastructure.
I like product engineering where the work touches real people and real constraints: cashiers, operators, managers, field hardware, unreliable networks, messy documents, background jobs, and systems that need to stay understandable when something fails.
| Area | What I care about |
|---|---|
| Operational products | Interfaces that help each role do the right next action, not generic dashboards that dump data |
| Retail and POS systems | Offline checkout, sync queues, receipts, barcode scanners, weight scales, orders, inventory, and payments |
| AI-assisted workflows | Model extraction where ambiguity exists, deterministic rules where business truth matters |
| Backend platforms | Typed APIs, generated SDKs, event logs, background workers, reliable state transitions |
| Infrastructure | CI/CD, Docker, PM2, Ansible, health checks, deployment scripts, monitoring, and recovery paths |
I am building a public writing repo around engineering lessons from private production work:
Topics include:
- role-centered UX for big operational systems
- human review as typed state
- AI workflow boundaries
- offline POS queue ownership
- reliability beyond
/health - shadow-run migrations
- rollback judgment
The repo uses synthetic examples only: no private screenshots, customer data, hostnames, credentials, raw payloads, or internal system names.
The pattern I keep coming back to:
- design around roles, not a vague generic user
- make the next safe action obvious
- keep model calls where ambiguity is real
- prefer deterministic code when the input is structured
- turn human review into typed, auditable state
- make degraded states visible instead of hiding them behind fallbacks
- treat generated SDKs and APIs as product surface
- delete obsolete paths after parity is proven
| System | Signal |
|---|---|
| POS ecosystem | Flutter mobile, Nuxt dashboard, Node/PostgreSQL backend, generated SDKs, real-time sync, offline flows, receipt printers, barcode and scale integrations |
| AI document workflow | OCR/model extraction, human review gates, typed snapshots, event logs, background workers, operator dashboards |
| Multi-tenant business platform | WebAuthn/passkeys, TOTP, JWT auth, real-time inventory, Flutter clients, deployment automation |
| Public field notes | Writing and visual assets that turn private engineering lessons into public-safe case studies |
Backend: TypeScript, Node.js, Express, Python, FastAPI, PostgreSQL, MongoDB, Prisma
Frontend: Vue, Nuxt, Tailwind CSS, operational dashboards, workflow-heavy interfaces
Mobile and edge: Flutter, Dart, offline-first state, printers, scanners, scales
Realtime: Socket.IO, SSE, sync loops, background workers
Infrastructure: Docker, GitHub Actions, PM2, Ansible, Azure, health checks, deployment scripts
AI workflows: OCR/model pipelines, structured extraction, human review gates, deterministic post-processing
