AI Engineer | PhD Candidate in AI & Multimodal Content Authenticity
I build production automation systems for photonics and semiconductor manufacturing at ficonTEC Service GmbH (since 2025) while pursuing my PhD at Technical University of Munich (TUM) and OTH Regensburg, focusing on robust detection of AI-generated images and text using multimodal deep learning and explainable AI.
- Production-grade photonics assembly automation (FAU-to-PIC alignment, laser soldering, wafer prober systems, SECS/GEM integration)
- Multimodal AI content detection (CNNs + ViTs for images, BERT-family + stylometry for text)
- Explainable AI (Grad-CAM, LIME, RISE, SHAP) for high-stakes visual and textual forensics
- ficonTEC → Delivered end-to-end customer systems with sub-micrometer precision and full factory integration
- BMW → 95%+ accurate road-induced vibration prediction models → ~30% reduction in physical prototype testing
- Boehringer Ingelheim → Automated RWE pipelines + R Shiny dashboards → 40% faster clinical data processing
- Ph.D. Candidate – AI & Multimodal Content Authenticity (TUM & OTH Regensburg, since 01/2025)
- M.Eng. AI for Smart Sensors & Actuators – Technische Hochschule Deggendorf (2024)
- B.Eng. Electronics & Communications – Mansoura University (2022)
Languages: Python • R • SQL • MATLAB
ML/DL: PyTorch • TensorFlow/Keras • scikit-learn • Hugging Face
Computer Vision: OpenCV • ResNet • EfficientNet • ViT • YOLO
NLP: BERT • RoBERTa • DeBERTa • stylometry
XAI: Grad-CAM • LIME • RISE • SHAP
Automation: State-machine PCM sequences • Multi-axis motion control • Vision metrology • SECS/GEM
Others: Docker • FastAPI • Power BI • AWS
- Email: Mohamed.Mady@gmx.de
- LinkedIn: linkedin.com/in/mohamedmady19
⭐ Feel free to explore my repositories below — especially the ones related to AI-generated content detection and my PhD work!