When Latifah Al Essa, founder and CEO of Ayadi and Nuur, speaks about mental health technology, the conversation quickly moves beyond platforms and products. It becomes about language, culture, memory and the invisible gaps between people and the systems designed to support them.

What emerges is not simply a discussion about digital therapy or AI chatbots. Instead, it is a broader conversation about what happens when entire populations are technologically underserved, not because the tools do not exist, but because the systems behind them were never designed to understand how those populations communicate emotion, distress and vulnerability in the first place.

That realization shaped the journey of both Ayadi and Nuur.

And increasingly, it is shaping a larger vision for culturally adaptive AI built specifically for the Arab world and its diaspora communities.

Beyond access

Al Essa launched Ayadi in 2021 at a time when the most visible challenge in the region was access to mental-health support.

Her background in cognitive rehabilitation, combined with field experience across Kuwait, Syria and the UK, exposed her to a recurring problem. Millions of Arabic speakers were struggling to seek support because of stigma, affordability barriers, limited mental-health infrastructure and a shortage of culturally aligned care.

Ayadi was created to respond directly to those realities.

The platform connected users with a regional network of 50 bilingual specialists while offering appointments within 24 hours through a privacy-focused digital platform designed around cultural understanding and accessibility.

However, as the platform evolved, Al Essa’s thinking evolved alongside it.

Over time, she began to realize that access to care was only one dimension of the challenge.

As generative AI rapidly entered the wellbeing sector, Al Essa observed that most emerging systems were being trained primarily on English-language datasets and Western behavioral assumptions. Arabic existed within many of these systems merely as a translation layer rather than as a deeply understood linguistic and cultural framework.

“These systems could technically translate Arabic,” she explains, “but they couldn’t truly interpret how emotional distress is communicated across Arab cultures, dialects or social norms.”

That gap became the foundation for Nuur, an Arabic-first emotional-intelligence platform powered by AI.

Why existing AI systems fail to understand Arabic users

For Al Essa, the issue is far more complex than language support alone.

Mental-health expression, she says, is deeply cultural.

In many Arab societies, emotional distress is rarely communicated through direct clinical language. People often express psychological strain indirectly – through physical symptoms, religious references, humor, social withdrawal or shifts in behavior rather than explicitly stating that they are anxious or depressed.

Most global AI systems fail to recognize those signals because they were never trained on those communication patterns.

At the same time, Arabic itself presents one of the most technically challenging linguistic environments for large language models.

Arabic is spoken by more than 400 million people across more than 30 dialects, many of which differ substantially in vocabulary and social context.

“The same emotional state will be expressed differently across these contexts,” says Al Essa. “A system that misses the dialect or the register often misses the person entirely.”

She argues that what ultimately breaks down in these systems is trust. And in the mental-health category, trust is foundational.

“If a user feels misunderstood, judged or mechanically replied to, they disengage immediately,” she explains. “In highly stigmatized environments, where opening up is already difficult, a single mishandled interaction can cost not just the user, but their belief that this kind of support can work for them at all.”

That is why Al Essa does not consider culturally adaptive AI to be a simple localization layer added onto existing systems. Instead, she describes it as an entirely different AI design challenge.

Building Nuur around memory and behavioral intelligence

Unlike many existing mental health applications, Nuur was intentionally designed outside traditional categories.

It is neither a therapist marketplace nor a generic wellness chatbot.

Instead, Nuur was built as an Arabic-first, youth-oriented emotional-intelligence platform centered around long-term memory, behavioral personalization and adaptive conversational intelligence.

The system combines conversational AI with journaling, mood tracking and assessments. Over time, it builds a behavioral profile that adapts tone, pacing and emotional interpretation according to long-term interaction patterns.

Al Essa often describes Nuur as “a digital twin of yourself in your pocket.”

That architecture is heavily influenced by her background in cognitive rehabilitation.

“In rehabilitation work, you learn very quickly that a person’s emotional baseline is not fixed,” she explains, arguing that traditional static assessments fail to capture the complexity of human emotional states.

As a result, Nuur focuses heavily on dynamic personalization and long-term conversational memory.

This creates an emotional-intelligence layer that evolves alongside the user rather than relying on rigid therapeutic flows or generic prompt-response systems.

The scale of that learning infrastructure is already significant.

To date, the platform has handled more than 16.2 million conversations, creating one of the region’s largest proprietary datasets focused on Arabic emotional interaction patterns.

For Al Essa, this dataset represents far more than operational growth.

It forms the foundation for culturally specific behavioral learning that thinner datasets simply cannot support.

The next frontier of AI wellbeing

Al Essa believes the future of AI wellbeing will depend less on increasingly larger models and more on emotionally intelligent infrastructure capable of cultural adaptation.

At present, she argues, most large language models are optimized primarily for productivity, search or general conversation. Very few are being designed specifically around culturally adaptive emotional intelligence, particularly for non-English populations.

That, she believes, is where the long-term opportunity for Nuur lies.

At the same time, she does not see Nuur as replacing human care.

Instead, she positions Ayadi and Nuur as complementary layers within a broader ecosystem.

Ayadi serves as the human-care layer, providing access to therapists and a trusted entry point into mental-health care.

Nuur functions as the AI-powered emotional-intelligence layer operating alongside it, offering daily support, personalization, memory systems and accessible engagement for individuals who may not yet feel ready for therapy or who require support between human touchpoints.

Together, the two platforms address a much broader spectrum of emotional wellbeing than either could independently.

Expanding beyond the Middle East

Al Essa sees significant potential for emotionally intelligent AI technologies across healthcare, education, workforce wellbeing and youth-support systems.

Her current focus includes advancing multilingual conversational intelligence, memory systems, and culturally adaptive behavioral models.

To support that expansion, the company has already raised more than USD 3 million from investors.

The UK has emerged as a particularly strategic market for the next phase of growth.

Al Essa says the country is home to one of Europe’s largest Arab communities and has built one of the world’s strongest ecosystems for healthcare innovation and responsible AI research.

Importantly, many Arabic-speaking communities in the UK continue to experience the same cultural and linguistic gaps in mental-health support that originally inspired Nuur’s creation.

“What I am bringing across is something the market doesn’t currently have,” she says. “An AI system designed around the cultural and linguistic realities of how Arabic speakers communicate about their mental health.”

Al Essa’s goal is to build the foundations of an AI ecosystem shaped around how Arabic speakers communicate emotion, identity and behaviour – one that is designed for the Arab world and the millions who speak the language far beyond it.

Related post

View all