Skip to content

We build AI that runs in production

Intelligent systems.Precision in action.

We turn the systems your business already runs into intelligent platforms, with AI integration, consulting, and evaluation you can measure.

In production today: Mehrab, our Quran learning platform, used by 30,000+ learners in 100+ countries.

While others build on the ground, we build above it.

Zad Code is a software house that works where AI meets real operations. We don't chase trends or sell one-size platforms. We study how your business runs today, then add intelligence where it makes a measurable difference.

Zad (زاد) is Arabic for the provisions you carry on a long journey. That is how we think about software: it should carry you further.

What we hold to

Deep expertise
Engineers who have shipped AI to real users, not just demos.
Precision
Clear scope, measured results, no guesswork.
Flexibility
We fit the tools and teams you already have.
Practicality
AI only where it pays for itself.

What we do

Four ways we can help, on their own or together.

  • AI integration

    Connect language, speech and vision models to the tools you already run, from CRMs and ERPs to support desks and internal apps.

  • AI consulting

    Find where AI actually pays off, choose the right models, and plan an architecture your team can own.

  • Evaluation and analytics

    Measure model quality against benchmarks built from your own data, and turn operational data into dashboards people use.

  • Product engineering

    Design and ship complete web, mobile and desktop products, from the first prototype to the app stores.

Case study

How we built Mehrab for 30,000+ Quran learners

Mehrab

محراب القرآن

Practice with AI. Perfect with a live teacher.

Mehrab's tajweed check: a learner recites a verse and gets a score with each mistake explained
Mehrab's memorization test: Al-Fatiha appears word by word as the learner recites it
A live one-to-one recitation lesson with a teacher on Mehrab

The challenge

Most Muslims have no reliable way to reach a qualified Quran teacher who holds an ijazah, and where those teachers exist, lessons are expensive. What's scarce is a qualified teacher, at a fair price, at the hour a student is actually free.

What we built

A live call with a verified teacher within seconds, free and unlimited, in 12 languages. When every teacher is busy, our own speech-recognition models take over: they listen to the recitation, test memorization and correct tajweed, so learning doesn't stop at 3am. It all runs on real-time infrastructure we host ourselves, on iOS, Android, Huawei, Windows and the web.

Inside the system

  • Word-by-word memorization tests

    Our own real-time Quran speech recognition follows a recitation and reveals each word as it is said.

  • Recitation and tajweed feedback

    Pronunciation and tajweed mistakes are detected, so learners know exactly what to fix.

  • Lessons that leave notes

    Recorded sessions are transcribed and turned into feedback students can review later.

  • Real teachers, around the clock

    Free voice and video calls with certified teachers, whenever a learner is ready.

Results

registered learners
30K+
countries
100+
hours of live teaching
15K+
app stores, plus the web
4

What this means for your business

Everything that runs Mehrab is what we bring to your systems: speech and language models shaped around your domain, evaluation against real usage, and infrastructure that stays fast and affordable as you grow.

Start a project

Who you'll work with

Portrait of Ahmed Hany

Ahmed Hany

Founder & CEO

Ahmed leads Zad Code and built Mehrab's platform across five operating systems, from live calling to the AI that listens to recitation. Before Zad Code he led engineering teams shipping web and mobile products, and built egytrains, which grew to about a million visitors a month with no paid marketing.

Connect on LinkedIn

How we work

Every engagement follows the same four steps, sized to the problem.

  1. Understand

    We map how your business runs today: the data, the workflows, and the decisions that matter most.

    You receive: a written assessment of where AI will pay off, and where it won't.

  2. Design

    We choose where AI earns its place and design a system that fits what you already have.

    You receive: an architecture and an evaluation plan, with the metrics we'll be judged by.

  3. Build and integrate

    We ship working software in short cycles, connected to your real tools and data.

    You receive: working software on your own data, connected to your tools.

  4. Evaluate and improve

    We measure quality against clear benchmarks and keep improving after launch.

    You receive: benchmark reports and measured improvements after launch.

Have a system that should be smarter?

Tell us what you run today and where it holds you back.