Our work: software built in Albania.

What we have built and where each piece of work stands today. No inflated numbers.

A laptop with verto’s revenue-by-product report and a phone with the fanar app
Real verto screens with demo data. The product is in Albanian.
Pilot completed

Verto Business in hospitality

The first full installation of Verto Business, piloted at a hospitality business in 2026 (the pilot has ended): table-side POS, a kitchen screen, stock that moves by itself and double-entry accounting, in one system.

What we built
POS, inventory, accounting, staff with roles, reports
Where it ran
On an on-site computer at the client
Status
The pilot has ended. What we learned went into the Hospitality & Restaurant package. Integrated fiscalization is in testing, not active.
Verto's quick POS, with demo data
Real verto screens with demo data. The product is in Albanian.
Two phones with the fanar app: the start screen and the map while driving
Real screens from fanar, an app in Albanian.
Live

Fanar, road alerts for drivers

A free app, in Albanian, where drivers warn each other about potholes, accidents, landslides and other hazards on the road. It opens straight in the browser and installs on a phone without an app store.

What we built
Map, one-tap reporting, on-road alerts
Where it runs
On Cloudflare, open to everyone
Status
Live and free. We are adding voice alerts.
Read-only phase

Ads system for clinics

A system that reads ad results across several accounts, links them to the leads that came in and proposes changes. In this phase it only reads and proposes: a person approves every spend, and no health data is sent to ad platforms.

What we built
Reading ad results, copy checks, reports
The rule
Not one lek without human approval
Status
Read-only. Automatic actions come later, with approval.
A phone showing a new contact from an ad and its path into the CRM
Illustrative example.
A laptop showing the state of our two servers and the services they keep running
Illustrative example.
Running

Our AI servers, NVIDIA GB10

We have built our own infrastructure on two NVIDIA GB10 servers. They run the language models we use, our internal services and some client systems. When a task can be done there, the data doesn't leave; when a service needs an outside provider (for example voice, or ads on Meta), we say so clearly.

What we built
AI models on our own servers, a private network, services
Why
Control over data and predictable costs

Have a problem that looks similar?

Write to us. We'll tell you honestly whether we can solve it and how.

+355 68 640 7594contact@zedigital.tech
Or write to us here

The message is sent from your own WhatsApp. This page does not store your details.