How I Built HeyYusuf: MSA, Gulf & Egyptian Arabic

I wanted to build a practical Arabic-learning app: something that would help a beginner move from unfamiliar words to a useful conversation. But before I could decide what an early lesson should contain, I had to answer a more basic question. Which Arabic was I helping someone learn?

That question shaped HeyYusuf. It affected the curriculum, the audio, the lesson sequence and the way I presented choices on a phone. Building the app meant working through those decisions together, then revisiting them as the learning path became clearer.

Why I built HeyYusuf

The experience I wanted to build had a straightforward purpose: help someone understand a little Arabic and begin using it. Practical vocabulary, listening, pronunciation and guided conversations would all contribute to that progression.

What interested me as a builder was the connection between those parts. A vocabulary exercise can make sense on its own, yet still leave someone unsure what to do when they hear a question. A conversation can be useful in principle, yet ask too much of someone who has only just started. The work was in making one activity prepare the learner for the next.

My role covered both the product and its curriculum architecture. That made lesson design part of building the app itself. Deciding what a beginner needed to know was also deciding what the interface needed to show, what audio belonged there and how much to ask at once.

The problem with treating Arabic as one learning path

One of the biggest problems I saw was the assumption that “learning Arabic” describes a single goal. Someone interested in reading formal material may need a different starting point from someone who wants everyday conversations in the Gulf or with Egyptian speakers.

Modern Standard Arabic, Gulf Arabic and Egyptian Arabic overlap, but they are not interchangeable learning experiences. If an app moves between them without making that clear, a beginner has another problem to solve: working out why the vocabulary or pronunciation seems to have changed.

I wanted that responsibility to sit with the product. A learner should know what they are practising and why it fits their goal. They should not need an understanding of Arabic varieties just to make sense of an introductory lesson.

For readers still choosing a learning tool, the Arabic-learning app comparison on HeyLanguages looks at that choice from the learner's side. The question behind my own product decisions was narrower: how could I make the chosen variety stay clear throughout the experience?

Why I separated MSA, Gulf and Egyptian Arabic

I built HeyYusuf around three separate paths. Each gives the learner a clear context for the material they are about to practise.

• The Modern Standard Arabic path is for learners whose goals include formal Arabic, reading and structured language foundations.
• The Gulf Arabic path focuses on learners interested in everyday communication in the Gulf.
• The Egyptian Arabic path gives learners a separate route into Egyptian speech and conversation.

This choice made the curriculum more demanding to organise. A shared topic does not automatically mean a shared lesson. The intention might be the same across paths, but the vocabulary, pronunciation and dialogue still need to belong to the variety being taught.

The distinction also had to survive beyond the path-selection screen. Naming three paths would mean little if the lessons then blurred them together. I treated dialect separation as a constraint on the whole learning experience, including the relationship between the text a learner sees and the audio they hear.

Designing the learning experience for beginners

Beginner-friendly progression became one of the main areas of iteration. I simplified early lessons and put more emphasis on getting learners into useful conversations.

The question I kept coming back to was how much someone needed before they could take part. Every additional word or explanation may be useful eventually, but it also adds something to remember now. Early lessons needed a manageable purpose.

I wanted practical vocabulary to have somewhere to go. A small set of words becomes more meaningful when it helps someone follow a short exchange. That gave me a way to think about sequencing: introduce enough to make the next activity approachable, then build on it.

On mobile, that same discipline applies to the screen. Instructions, learning content, playback and the next action compete for a limited amount of space and attention. Mobile UX was part of the teaching problem: the learner was already doing something difficult without also having to interpret an overly busy interface.

Building pronunciation and conversation into the product

Pronunciation and listening belonged inside the learning path. Seeing a word, recognising it when spoken and trying to say it are related tasks, but one does not automatically give you the others.

Audio quality mattered for that reason. For a beginner, audio is a reference for something they may never have heard before. Clarity and consistency affect whether the exercise is usable, especially when the learner is also trying to understand a new phrase.

Guided conversations gave the vocabulary a context. I wanted the learner to have enough support to follow an exchange without needing to invent every response from scratch. That meant thinking about what the conversation assumed they already knew, and whether the preceding material had actually prepared them.

Keeping those pieces aligned was a substantial part of the build. A conversation, its supporting vocabulary and its pronunciation practice needed to work as one sequence within the selected Arabic path.

What changed during development

The learning path went through multiple iterations. The main shift was towards simpler early lessons and a stronger connection to useful conversations.

That changed how I judged the material. I became less interested in how much a lesson could contain and more interested in what it enabled the learner to do next. If the next conversation required too much unfamiliar language, the sequence still needed work, even if each individual exercise seemed reasonable.

Simplifying also meant making choices about timing. Material can be useful and still arrive too early. Moving it later can give the learner room to understand the current step without removing it from the wider curriculum.

These were product changes as much as content changes. The shape of a lesson affects its pacing, its screens and the amount of explanation the interface has to carry. Revising the learning path meant reconsidering the experience around it.

The hardest product decisions

The hardest decisions involved things that depended on each other.

Curriculum architecture and dialect separation. Three paths needed a coherent overall structure while leaving room for differences in the language itself. A tidy structure is only helpful if it supports the lesson being taught. I had to keep both concerns in view.

Simplicity and enough context. I wanted beginners to start using the language early, but a short lesson still needs to prepare them properly. Removing an explanation is only helpful if the activity remains understandable. The challenge was deciding what belonged at that particular point in the path.

Character consistency. The characters were another part of the experience that needed to hold together as the product evolved. Consistency matters across lessons: visual changes should not distract from the language someone is trying to learn. That work sat alongside curriculum and interface decisions, rather than being a final decorative pass.

Audio quality and mobile UX. Listening exercises depend on both the material and how it is presented. I had to consider the spoken content alongside the screen supporting it. Neither could be judged entirely in isolation.

These decisions made the build interesting. Much of the difficulty came from keeping several parts of the product coherent while changing one of them.

Launching the first version

HeyYusuf is now available on Google Play. iOS is coming next.

Getting the app to this point meant bringing the paths, lesson progression, audio, characters and mobile experience together into something a beginner could use. The launch is a milestone in that work. It does not settle every question about how the learning experience should develop.

Shipping the first version gave me something real to test and improve. The next phase is learning from how people actually use it.

What I learned

The clearest lesson was that the way I defined the learning problem shaped almost every other decision. Once I committed to separate Arabic paths, that choice carried through the curriculum, pronunciation, conversations and interface. It could not stay a positioning statement on the opening screen.

I also learned how much judgement sits inside the word “simple.” Making an early lesson smaller requires deciding what the learner can reasonably do without, what support they still need and where the omitted material will belong later. It is careful sequencing work.

Finally, working across product, design, development and curriculum made their dependencies very visible. A lesson revision can change the audio it needs. A conversation can expose a gap earlier in the path. The interface can make an otherwise clear activity harder to understand. I needed to consider those connections when deciding whether a change actually improved the experience.

Where HeyYusuf goes next

iOS is the next platform step. Beyond that, I want to keep refining the progression from early vocabulary to guided conversation, while preserving the clarity of the three paths.

The question I want to keep asking is specific: does this step give a beginner enough understanding to attempt the next one? That is a useful standard for deciding what to simplify, what to explain and what to build next.