UAE CNTXT opens office in USA to support US government agencies

CNTXT AI, a UAE-based data and AI company, has reached more than $300 million in committed revenues, most of it from U.S. hyperscaler, AI developers and large enterprises in the US, and has expanded operations to the United States with its first office in the San Francisco Bay Area. CNTXT AI’s new office will support American government agencies, technology companies, AI developers and large enterprises that need access to sovereign AI infrastructure and expertise delivered inside the Middle East.

Founded in 2024, CNTXT AI spent two years developing its Arabic-first AI models, building its data operations and supporting regional deployments. In June 2026, the company raised a $60 million Series A, acquired Actualize, an enterprise AI company building dialect-aware Arabic voice agents, and launched three commercial offerings. CNTXT AI currently supports more than 250 government and enterprise customers across three core areas:

CNTXT AI supplies the data that hyperscalers, foundation model developers and enterprise AI teams train and evaluate on: sourcing, multimodal annotation, expert RLHF, model evaluation and red teaming.

It specialies in hard-to-source data: scientific, technical, industrial and physical-world, where accuracy and context decide whether a model works. All of it runs on CNTXT AI’s own platform, with automated pipelines, quality controls and model-assisted pre-annotation, so the work moves fast without losing accuracy, and domain experts on every judgement call. Its Arabic and MENA data helps global AI leaders build models that work in underrepresented languages and real-world settings.

In terms of AI models, the best known is Munsit, the most accurate Arabic voice AI, ranking #1 on Hugging Face’s Arabic speech model leaderboard. Its speech-to-text and text-to-speech models are trained on more than 110,000 hours of authentic Arabic-language audio across more than 25 dialects. Munsit has more than 300,000 users, and U.S.-based companies using it include Boston AI, SPSoft and Clarity. They use it to serve Arabic-speaking customers in products built in English, across customer service, voice agents, and transcription.

Additionally, CNTXT AI has dedicated robotics training facilities and a robust data collection network for Physical AI model training, capturing real-world data across more than 300 use cases. It continuously generates training data daily from diverse real-world environments, with consent and chain-of-custody documentation to meet rigorous privacy standards. CNTXT AI has also filed a U.S. patent application for its approach to prevent “diversity collapse,” where robots successfully copy a single training demonstration but fail when actual real-world conditions vary.

Physical AI systems need large volumes of diverse, task-specific real-world data that simulation alone cannot produce. CNTXT AI’s platform captures real work in live environments, verifies it through model-assisted labeling and human review, and delivers it as structured training data that robotics and AI systems can ingest directly.

“We built CNTXT AI around the hardest part of AI, providing real-world data at the quality and scale models actually need, which is why customers keep choosing us over others,” said MAS, Founder and CEO of CNTXT AI. “What comes next is bigger: whoever can capture real work, legally and at scale, decides which AI makes it out of the lab and which stays a demo. We intend to be the reason it happens.”

CNTXT AI’s growth is supported by deployments and partnerships across major regional organizations, including the Technology Innovation Institute (TII), Nationwide Middle East Properties and Venture One. The company’s work includes specialized linguistic datasets, multilingual annotation frameworks and evaluation methodologies supporting AI research and Arabic centric language model architectures.

MAS continued, “Conducting business successfully in the Middle East requires a deep understanding of Arabian business culture, compliance mandates, enterprise environments, dialects and other requirements that directly affect adoption. Our expansion to the bay area gives U.S. organizations a direct path into those capabilities, and into markets across the Middle East where sovereignty, regulation, language and local operating requirements shape every major AI deployment.”