Farida Mohamed

Completed 10 applied GenAI projects covering prompt engineering, RAG, and the Gemini API, including health and detection/classification use cases. Built and deployed project workflows across multiple applied AI tasks.

Completed Level 3 of the AI/ML certificate program emphasizing prompt engineering, RAG, and generative AI. Worked with Gemini API across 10 applied projects.

Designed and compared generative models to synthesize realistic Arabic handwritten text conditioned on input text. Implemented and evaluated CVAE, cGAN, and a transformer-based generator using PyTorch and HuggingFace.

Led implementation of a graduation project for an AI-powered IoT security system using multi-model fusion and Zero Trust architecture, including model training and deployment. Integrated a NIST SP 800-207-inspired Zero Trust policy engine and achieved 92.56% CVE severity classification accuracy and 91.72% network traffic threat classification accuracy.

Completed Level 2 of the AI/ML certificate program with coverage of NLP and speech processing topics. Also trained on LSTM/GRU sequence models and image processing with OpenCV.

Completed Level 1 of an AI/ML certificate program focused on core ML and foundational tooling. Covered Python, data processing, and fundamental neural network architectures.

Had a research paper accepted for IoTSec4CIS 2026 and presented related work during the June 2026 workshop on IoT security for critical infrastructure systems.

Completed an AI Department intensive training program at WE Telecom Egypt in April 2026. No additional technical scope was specified in the source.

Built a stateful, multi-node LangGraph workflow with conditional control flow and LLM integration for automated data analysis.

Built a live-camera pipeline combining OpenCV face detection with face_recognition embeddings to identify authorized users in real time. Added DeepFace emotion analysis and MediaPipe hand-landmark tracking to enable gesture-controlled slide navigation.

Analyzed 27,000+ hypermarket transactions across 5 branches using SQL, then cleaned and engineered features in Python (Pandas). Built an interactive Power BI dashboard highlighting verified 2024 revenue and key sales/per-store insights.

Designed a multimodal sensor fusion approach for detecting and tracking moving objects as part of an AI project.

Built a multilingual Arabic/English Retrieval-Augmented Generation (RAG) system to query museum artifact documentation. Added a speech-to-text recognition component for processing spoken input.

Benchmarked multiple resampling strategies for cybersecurity datasets and documented results with confusion matrices and G-mean comparisons. Reported best experimental design performance of 90.75% accuracy and 89.84% G-mean using Random Forest on UNSW-NB15 and KDD99.

Completed a data analysis training program (60 training hours) covering Excel, SQL, Python, and Power BI for dashboards and analytics reporting. Included training on data cleaning, queries/joins, and interactive KPI/report building.

Grade: A+

B.Sc. in Computer Science (Artificial Intelligence) from E-JUST, graduating in 2026 with final CGPA 2.62 and an A+ graduation project on an AI-powered IoT security system using multi-model fusion and Zero Trust.