CoreTechX launches first end-to-end on-premise AI system for Arabic handwriting

The system scans handwritten pages, transcribes text, structures content into searchable data, and enables integration with existing digital systems

  • PUBLISHED: Wed 25 Mar 2026, 2:57 PM

CoreTechX has launched a fully on-premise AI platform designed to help government ministries, courts, and archival institutions digitise handwritten Arabic records without compromising data sovereignty.

The system operates entirely within institutional infrastructure, allowing organisations to process, transcribe, structure, and search handwritten Arabic documents without routing data to external servers or cloud providers.

"We did not rely on off-the-shelf OCR," said Fahad Durukan, co-founder of CoreTechX. "We built our own end-to-end pipeline for Arabic handwriting to ensure the system understands the grammar and historical context of the script."

Arabic handwriting poses distinct challenges for AI systems. Cursive flow, diacritics, ligatures, and regional and historical variation in writing styles make automated recognition complex. Standard vision and language models perform poorly on such material.

CoreTechX's proprietary OCR system, introduced earlier this year, recorded a Character Error Rate of 3.6 per cent on the modern Khatt dataset and 6.3 per cent on the Muharaf historical manuscript dataset - results the company says set a new benchmark for Arabic script recognition.

The newly launched platform embeds that capability into a secure local environment built for institutions subject to compliance requirements that prohibit third-party API use.

"Arabic handwriting is one of the hardest challenges in document intelligence due to its cursive structure and wide variation across eras. Arabic content deserves first-class technology built specifically for it," said Fahad Faisal Fahad AlSaud, co-founder of the company.

The system scans handwritten pages, transcribes text, structures content into searchable data, and enables integration with existing digital systems. Courts can retrieve case histories; ministries can analyse historical records; archival institutions can make manuscripts accessible to researchers.

"For too long, important decisions were being made without access to decades of data … because it was locked in a format machines cannot understand," AlSaud said. "This is not just digitisation; it is about unlocking the past to help build a more informed future."

CoreTechX describes the launch as a shift from backend technology provider to a comprehensive document intelligence platform for real-world institutional use.