I work on both ends of the same question: how a system can understand, evaluate, and support human reasoning — as a researcher, and as a founder.
PhD in Computer Science, specialized in AI applied to learning and in RAG / LLM architectures. My work moves from published research to products running in production.
Design of a cheating detection system for online exams, combining computer vision and AI models. The interactive proctoring tool resulting from this project was published in Neural Computing and Applications (2026).
View publicationIntelligent exam platform built on generative AI, designed together with faculty to integrate directly into existing teaching workflows.
Combining process mining and clustering to understand how students solve programming problems, and to automatically detect their strategies.
ITS 2023 CSEDU 2022Comparative study of automated program graders in a MOOC context, benchmarking automated assessment against human grading.
Computer Applications in Engineering EducationDesign and deployment of a learning analytics dashboard, integrating multi-source data in a production environment.
Multi-tenant SaaS platform built on a RAG architecture: document ingestion, embeddings, vector search, reranking, and multilingual QA. Backend architecture on FastAPI / Docker / Qdrant / PostgreSQL, deployed in production.
kestya.comA learning platform for Arabic, focused on dictation and written expression.
abjedi.com