In a region where global tech giants often overlook linguistic nuances, one startup dared to tackle a massive market gap – the Arabic-speaking world. Founded by Abdullah Asiri, an AI enthusiast since his college robotics days, Lucidya recently achieved a massive milestone, closing a $30 million Series C funding round (CSB)—the largest AI round ever recorded in the MENA region.
This achievement wasn’t an overnight success. It was the culmination of years of relentless R&D, strategic pivots, and learning hard lessons from past entrepreneurial attempts. Here is the story of Lucidya’s journey, the major problem it solved for 400 million Arabs, and how can MENA entrepreneurs learn from this experience ! Enjoy 🙂
The Founder’s Blueprint – Early Failures and an Obsession with AI
Abdullah’s journey into entrepreneurship was deeply rooted in his passion for Artificial Intelligence. He studied Computer Engineering at Virginia Tech, where he headed the Robotics Club and competed nationally. He later pursued a Master’s degree in Computer Science focused on AI at KAUST.
This deep-seated passion led him to launch his first AI company in 2011, ShopMate, which used AI and geofencing technology to send personalized shopping notifications to users passing by specific stores. Though innovative, the venture failed by 2013. Abdullah attributes this failure to critical mistakes that served as valuable lessons for his later ventures:
- Poor Timing (Too Early) – The market was not ready for the concept yet.
- Wrong Investors – He emphasized the importance of thorough background checks on investors and noted that bringing in the wrong investment partners was a major downfall.
- Flawed Business Model – Initially trying to charge both consumers (though the app was free) and brands proved problematic.
His second venture, Waqood Tech (started around 2013), was a bootstrapped studio focused on creating mobile apps during the mobile app economy boom. This venture taught him another crucial lesson – lack of focus kills opportunity. They attempted to launch several apps per year, but failed to commit the necessary concentration required for an app to succeed (which can take two or three years). The company also became distracted by service projects for customers.
By 2016, Abdullah had opened and closed over 10 or 12 investment rounds across his previous startups. These experiences provided him with the necessary resilience and knowledge for his next, most successful undertaking: Lucidya.
Identifying the Gap Why 400 Million Voices Were Silenced
The foundation of Lucidya came from observing a staggering market failure involving Arabic-speaking customers.
Around 2016, Abdullah’s co-founder, Hatem Kamili (who ran a marketing agency), approached him with a core technical question – Why were there no software tools capable of classifying social media posts and comments in the Arabic language?
At the time, social media listening software – tools that analyze sentiment (positive/negative/neutral) about a brand – were globally booming. However, global software vendors completely ignored the Arabic language; their AI simply could not comprehend Arabic.
“There was no tool in the world that even thought about serving the Arab person… We are talking about 400 million Arabs in the world. Finally, there was a chance for me to use my passion and my enthusiasm to solve a problem – which is AI”.
The immediate consequence was disastrous for businesses in the region. Companies were paying significant sums for global tools, resulting in “terrible, faulty analysis”. To compensate, marketing agencies and brands had to employ staff solely dedicated to manually reviewing and labeling thousands of comments as “positive” or “negative” (lol) – a task Abdullah called “torture” for human employees.
Abdullah realized he had found a massive market gap, a vacuum created by large, billion-dollar US and European companies that deemed the Arabic market too small or too difficult to justify the necessary R&D investment. He confidently stepped in, knowing that solving the Arabic language AI challenge was his “game”.
The Challenge of Building Arabic AI – R&D and World-Class Accuracy
Lucidya’s journey from 2014 (R&D phase) to its 2016 launch required overcoming two tremendous, interconnected hurdles that had deterred global competitors:
1. The Data Set Desert
The first major difficulty was the almost total lack of sufficient, high-quality annotated data sets in Arabic. Unlike English or Spanish, where research institutions produce massive data sets, the Arab world lacked this foundational research infrastructure focused on commercialization.
Furthermore, Arabic complexity is immense. The language includes many varied dialects – Egyptian, Saudi, Lebanese, Iraqi, and Moroccan – all with different writing styles, grammar, and meanings. The Arabic language is “very difficult” to build AI models for.
2. The Talent Scarcity
The second challenge was finding AI talent who were technical experts and deeply understood the nuances of the Arabic language and its dialects.
The Solution – Building a Data Factory
Lucidya spent nearly two years (2014-2016) in deep R&D. Leveraging the expertise of co-founder Dr. Zuhair Khayyat (who holds a PhD from KAUST), they focused on developing internal data sets.
Crucially, they launched a sophisticated, internal Crowd Annotation Portal. They hired 70 to 80 anonymous freelancers from across the Arab world. For every piece of text, three different people provided a sentiment label (positive, negative, neutral) and identified the dialect. This “voting” consensus system ensured data quality and prevented cheating.
This rigorous, iterative process allowed them to build a massive, high-quality data set. Their initial model accuracy was 72%. But through continuous trial and error, they rapidly improved (77%, then 84%, 87%). By 2020-2021, Lucidya reached 92% accuracy – the highest rate in the world for Arabic sentiment analysis, covering 12 dialects.
The Smart Approach to B2B Sales
Lucidya’s initial target clients were large enterprises and government entities. Selling to these demanding clients as an unknown startup with an AI product (in 2016, when most people didn’t even understand what AI meant) was exceptionally difficult.
Their very first client came aboard despite the product being “very, very bad” (sometimes failing to work or log in). This sale was secured because the client trusted Abdullah’s personal reputation and previous track record, believing he could develop the product, even if the Minimum Viable Product (MVP) wasn’t yet strong.
As they scaled, the usual business-to-consumer (B2C) “word-of-mouth” strategy was too slow for the enterprise sector. Lucidya needed a clever strategy to gain notoriety.
The Industry Report Strategy
Lucidya leveraged its access to public social media data to grade the performance of entire industries. They published reports evaluating, for example, the 10 largest banks in the Gulf, ranking them based on public sentiment, response times, most common complaints, and overall positive vs. negative volume.
These reports, which were costly to produce as Lucidya paid for every piece of data analyzed, served as powerful, targeted marketing. Brands started following Lucidya because they wanted to know where they ranked, turning the company into an essential source of market insight.
This approach sometimes “backfired”. Abdullah recounted a meeting where an executive threw the report onto the table, furious that Lucidya had exposed their low performance publicly. However, these tense meetings usually ended with the brand subscribing to Lucidya’s services to gain access to the raw data and solutions needed to improve their reputation.
The AI Market Reset – From Listening to Agents
Today, Lucidya has evolved beyond simple social listening. It is now a Customer Experience (CX) company, helping brands improve the entire customer journey, from marketing campaigns to post-sale support.
The emergence of Large Language Models (LLMs) and Generative AI (GenAI) created a massive shift—what Abdullah calls a “Market Reset”. While LLMs lowered the barrier of entry for competitors, they significantly benefited Lucidya:
1. Market Validation – LLMs showed the world the power of AI, turning AI tools from a “nice-to-have” into a “must-to-have” for every major corporation with an AI strategy, thereby expanding Lucidya’s total addressable market.
2. Product Expansion – LLMs allowed Lucidya to dramatically increase its product offering (now six products strong) and deliver greater value.
One notable pivot enabled by LLMs was in conversational AI. Lucidya was in discussions to acquire a large Egyptian chatbot company for a significant sum, but when LLMs matured, they stopped the acquisition and quickly built their own product, the Lucidya AI Agent, easily.
These new tools include:
• Lucy (Virtual Data Analyst): A GenAI-powered copilot for human analysts. Instead of manually navigating dashboards and filtering data, a client can ask Lucy questions like, “Where is the most important branch to open?” and Lucy immediately provides data-driven answers.
• Lucidya AI Agent (Conversational Agent): This goes far beyond basic chatbots. The Agent is transactional, meaning it completes complex, end-to-end tasks without human intervention. For example, the agent can handle a full refund process: verify the order, check internal policy APIs, communicate with the finance portal, get confirmation, and inform the customer – all automatically.
The Agentic World – A Vision for the Next Five Years
Looking ahead, Abdullah foresees a massive disruption in workflows, driven by the emergence of these AI Agents.
In the customer service field, he predicts that within five years, 90% or more of customer service interactions will be handled by AI Agents. Human customer support staff will transition from performing repetitive tasks (like answering the same questions or doing “copy-paste” replies) to becoming supervisors who review high-stakes operations or handle highly complex, rare “corner cases”.
He also sees a future where AI Agents communicate seamlessly internally within a company. For example, during peak sales seasons, a Customer Service AI Agent could communicate directly with an Operations AI Agent to resolve logistical issues.
This shift will have huge implications for entrepreneurship.
“The investment required to reach the output of a company with 100 employees, you can do that now with 10 employees only. This will increase the number of small and medium companies in the world”.
The “Market Reset” means companies must upskill their employees. Abdullah believes that soon, applicants will submit their CV alongside their personal AI Agent’s CV, highlighting how they leverage AI to boost their productivity and output.
What is in it for MENA Entrepreneurs ?
Drawing from his successes and failures, Abdullah offers several key insights for founders operating in the MENA region, especially those looking to leverage AI:
1. Don’t Start an AI Company, Start a Solution Company
The core mistake is trying to be “an AI company for the sake of being an AI company”. Lucidya succeeded because they identified a real, unsolved problem (the lack of Arabic language processing) and only then applied AI as the necessary solution. Entrepreneurs must validate whether their target market genuinely needs the product.
2. Validate Fast with an MVP
Abdullah’s first company built for over a year before launch, which he calls an error. Today, with the rapid development tools available through AI, you can build an MVP quickly to validate market need and ensure retention. The current environment is the “Best time to start” an AI company, as investors are keen and MVPs can be built with few resources.
3. Embrace Continuous and Fast Innovation
When competitors inevitably arrive (especially since LLMs lowered the barrier to entry), the absolute best competitive advantage is fast and continuous innovation (the “moat”). Abdullah states that Lucidya maintained its lead—even as LLMs lowered the complexity of Arabic processing—by rapidly building new adjacent features and products (like AI Agents and Lucy).
4. Deepen the Value Chain
Lucidya didn’t just stop at social listening. Once they understood what customers wanted (sentiment analysis), they moved to the next logical step: helping companies respond and act (Customer Service and AI Agents). Entrepreneurs should either dive deeper into solving a single problem in increasingly effective ways, or solve adjacent pains that naturally follow the first solution.
Lucidya’s journey is a powerful demonstration that necessity breeds innovation. By focusing on a profound linguistic gap ignored by global competitors, the company not only built a massive business but pioneered a world-class, data-driven solution tailored specifically to the diverse Arabic consumer, effectively giving 400 million people a voice in the global digital landscape. The key to future success, as always, is remaining fast and relentless in the face of change, using technology to lift the burden of repetitive tasks and allow humans to focus on the truly creative and deep thinking