DALIA ALKILANI
+963949546996 • daliaalkelani@gmail.com • Hama, Syrian Arab Republic • LinkedIn • GitHub
AI/NLP Engineer specializing in Arabic language models — fine-tuning AraBERT and MARBERT and deploying them into production environments. I build end-to-end ML systems across NLP and computer vision, from data preprocessing to model training and API deployment. I also mentor young makers in robotics and embedded systems, and create tech content that simplifies AI and inspires the next generation of innovators.
- Designed and built an embedded robotics prototype for pest detection using microcontrollers and environmental sensors.
- Integrated Arduino-based hardware with sensors and control components.
- Performed system-level debugging and optimized reliability during testing.
- Validated system performance through real-world testing and technical demonstrations.
- Implemented sensor integration and control systems using Arduino and external sensors for competition-ready robotics platforms.
- Mentored participants in embedded systems design and firmware troubleshooting to strengthen practical hardware and firmware skills.
- Delivered hands-on workshops on system integration, testing procedures, and competition readiness to improve team performance.
- Mentored teenagers in robotics fundamentals and competition preparation.
- Provided technical guidance and online support during robotics activities.
- Facilitated problem-solving discussions and assisted students with project challenges.
- Conducted virtual sessions to support learning and feedback.
- Contributed to a community initiative focused on robotics education and youth engagement.
- Constructed a labeled dataset of Arabic university inquiry texts through domain-focused collection, structured annotation, normalization, and preprocessing.
- Developed a multi-class NLP framework by fine-tuning AraBERT and MARBERT using Hugging Face Transformers for improved contextual understanding of academic queries.
- Designed a stratified 5-fold evaluation protocol to validate model stability across different data partitions.
- Exposed the trained system via a FastAPI service, enabling real-time inference in a production-oriented environment.
- Built a real-time bidirectional translation system between sign language and speech using Python.
- Implemented hand tracking using MediaPipe and OpenCV; trained Random Forest classifiers for gesture recognition (numbers, letters, words).
- Integrated speech-to-text and text-to-speech using speech recognition and gTTS.
- Developed a GUI using CustomTkinter with real-time interaction features.
- Built a full-stack peer-to-peer marketplace MVP using React and Tailwind CSS for the frontend and Laravel for backend services including authentication, APIs, and core marketplace logic.
- Implemented core listing and matching features in a lightweight web prototype to demonstrate feasibility and user flow.
Al-Wataniya Private University, B.Sc. in Computer Engineering, (GPA: 82% (3.09)), Oct 2021 – Jul 2026
- Relevant coursework: Natural Language Processing; Machine Learning; Embedded Systems.
- Led live robotics demonstrations showcasing sensor integration and control systems to engage attendees.
- Translated embedded system concepts into step-by-step troubleshooting explanations for event visitors.
- Produced short-form AI-generated video content and refined prompts to increase creative consistency.
- Aligned AI content and prompt strategies with marketing campaign objectives to improve audience fit and messaging cohesion.
- Arabic — Native
- English — Professional Working Proficiency