Skip to content

AI Lite World

Ammar Al-Najjar

AI Researcher & Machine Learning Engineer

I work on adversarial robustness and trustworthy machine learning — how models fail under attack, and what makes them hold. Alongside the research I build and ship AI-powered web and mobile products.

Projects

03

Resume Builder Build a professional CV in five languages, pick from four print templates and export a typeset PDF. AI-assisted writing, a career-scoring model, and analytics on how the document reads. ReactFastAPIPostgreSQLLLM
TestMe Turn a PowerPoint deck into a live, timed multiple-choice quiz that students join from a link or QR code — no student accounts, no app. ReactFastAPIPostgreSQLLLM
Learn Magyar Language Magyar is a free, self-paced online platform for learning the Hungarian language, offering clear and easy-to-follow lessons that allow learners to progress at their own pace. The platform supports learners from diverse linguistic backgrounds through multilingual instruction in **English, German, Hebrew, and Arabic**, making Hungarian language learning more accessible and inclusive. ReactFastAPIPostgreSQLLLM

Publications

04

2026
Breaking the trade-off between performance and cost with an improved untargeted APGD
Ammar Al-Najjar, Mark Jelasity

ESANN European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning
2026
A comprehensive analysis of objective functions in adversarial training
Ammar Al-Najjar

Pollack Periodica
2024
Adversarial training without hard labels
Ammar Al-Najjar, István Megyeri, Mark Jelasity

ESANN European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning
2022
PCA improves the adversarial robustness of neural networks
Ammar Al-Najjar, István Megyeri

ESANN European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning