Learning that adapts, research that understands Arabic
Adaptive learning and research, built for Arabic
Educational institutions and learners need tools that adapt to how a person actually studies, and research platforms that understand classical Arabic and religious or academic source material rather than just matching keywords.
Why a generic tutor does not work in Arabic
Two different problems get bundled together in edtech: how a learner is taught, and whether the system understands the material at all. Fekr addresses the first by adjusting to the learner's pace instead of pushing a fixed curriculum. The second is a language problem, and it is the one we build models for: classical Arabic, diacritics, and religious or academic source text handled at the source level rather than by keyword match.
Adapts to the learner, not the syllabus
Fekr tracks pace and weak points and adjusts the study plan, raising exercise difficulty as performance improves and explaining a mistake immediately rather than days later.
Sources understood, not matched
Semantic search over classical Arabic and academic source material returns passages that answer the question, and every generated passage maps back to where it came from.
Questions about data, answered plainly
0Plus lets a non-technical team ask about its own data in Arabic or English and get an answer with a chart and a written explanation, rather than a dashboard to go and interpret.
What institutions and learners get
- Adaptive learning
A study plan that responds to how the learner is actually doing.
Fekr is built around subject-specific tutors rather than one generic assistant, each shaped around its domain. The plan moves with the learner: exercises scale with performance, and a wrong answer gets an explanation in conversation rather than a grade days later.
Analytics & Learning - Research over Arabic sources
Cross-referencing that used to take a week of manual work.
Tbyaan searches the Quran, authenticated Hadith, Tafseer commentary and Islamic Q&A datasets together, understanding classical Arabic and religious context rather than matching keywords, and its research agent returns a structured, source-backed draft instead of a list of links.
Islamic Knowledge AI - Unambiguous Arabic text
Diacritics restored before the text reaches a learner or a speech engine.
Sadeed reads the grammar and meaning of the whole sentence to put Tashkeel back, which is what lets the same written word be read correctly in context. It is a preprocessing step for text-to-speech, search and language-learning tools, and it runs on a compact model rather than a large one.
Translation & Localization
Deployment options
Cloud, on-premises or dedicated in-Kingdom deployment, depending on what the institution's own data policy allows. The Trust Center sets out data residency, access control and the deployment options in full.
Let's talk about what you're trying to build.
Tell us the problem. We'll tell you honestly whether AI is the right answer.