How Tarjama is Bridging the Arabic Digital Divide

The Problem

Tarjama addresses the linguistic isolation of the Arabic-speaking world. Despite being the fifth most spoken language globally, Arabic is chronically underserved by high-quality digital content and artificial intelligence. Most global Large Language Models struggle with the twenty-two distinct regional dialects and the intricate cultural nuances of the Middle East, which creates a significant barrier to entry. Furthermore, the Middle East region has faced one of the world's lowest rates of female participation in the workforce. Many highly educated women have been excluded from the labor market due to rigid office structures or a lack of flexible, remote opportunities that align with their cultural and domestic realities.

Nour Al Hassan, Founder & CEO of Tarjama

The Solution

Tarjama has built a proprietary AI ecosystem known as Arabic.ai which is powered by its own Large Language Model called Pronoia. Unlike general-purpose models developed in Silicon Valley, Pronoia is trained on over sixteen years of professional, human-verified linguistic data comprising billions of words. This allows it to handle complex legal, medical, and governmental tasks with a level of cultural fluency that generic AI cannot match. To ensure accuracy and maintain a social impact mission, Tarjama employs a humans-in-the-loop model. This utilizes a vast network of thousands of linguists, roughly ninety percent of whom are women, who work via a flexible gig-economy platform to verify and refine AI outputs.

Business Model

The company generates recurring revenue by selling subscriptions to its AI platform, which allows large organizations and governments to automate their translation and content workflows. To ensure total accuracy for sensitive legal or medical documents, they offer a managed service where a network of thousands of female linguists verifies the AI's output. By using their proprietary model to handle the bulk of the work, they significantly reduce operational costs and pass those savings to their clients while maintaining high profit margins

Tarjama Management Team at Headquarters in Riyadh, Saudi Arabia

The Impact

Tarjama has demonstrated significant growth and social impact through its scale, technical performance, and workforce development. To date, the company has raised $20 million in total funding, including a recent $15 million Series A in 2025 to accelerate its Arabic-first AI infrastructure. While its "human-in-the-loop" model has automated 50% of business content with perfect accuracy. The company operates a network of over 200,000 vetted linguists, roughly 70% to 90% of whom are women, providing high-skilled employment in regions with historically low female labor participation. Over its 16-year history, Tarjama has processed more than 2 billion words across 55 languages and 22 Arabic dialects for over 700 enterprise clients. This operational efficiency has fueled a 20% compound annual growth rate in revenue over the last three years, proving the profitability of localized impact.

The Innovation

The true innovation lies in Tarjama’s "agentic" approach and its specialized dataset. While global tech giants rely on scraping the public internet, Tarjama uses its massive repository of professional translations to train models that are significantly more precise for enterprise use. Its proprietary Large Language Model, Pronoia, has achieved a 19% performance lead over global models like GPT-4o on specific Arabic linguistic benchmarks. Additionally, they have launched the Arabic.ai Academy, which focuses on upskilling hundreds of enterprise and government leaders in the region. This turns the enterprise from a service provider into an educational engine that prepares the local workforce for the AI era.

What Needs to Improve

To maintain its lead, Tarjama must continue to deepen its dialect support, particularly for the linguistically distinct Maghrebi dialects of North Africa which remain a challenge even for advanced models. As their proprietary AI becomes more efficient, the company faces the continuous challenge of upskilling its female workforce to transition from traditional translation into roles like AI prompt engineering and data auditing to prevent displacement. Finally, the enterprise must navigate the aggressive entry of global tech giants into the Middle Eastern market by proving that a "region-first" model offers security and accuracy advantages that generic global tools cannot replicate.

Sources:

https://www.arabnews.com/node/2631332/business-economy https://www.al-monitor.com/originals/2025/08/how-arabic-translation-firm-jordan-became-gulf-ai-player https://slator.com/tarjama-raises-usd-15m-in-push-for-native-arabic-ai-models/ https://waya.media/pronoia-powers-arabic-ai-translation-a-game-changer-or-industry-disruptor/