Amal EL Mabrouk

– Developed an automated image quality control system for internal algorithms using multimodal LLMs, including fine-tuning of LLAMA Vision. – Designed a pipeline for the automatic generation of visual rules, combining web crawling, image analysis, coherence scoring, and exception detection. – Automated the integration of results into Grand Shooting's specification system through API connections.

– Developed a content-based recommendation system using Flask, NLTK, Spacy, and Word2Vec to calculate the cosine similarity between courses. – Created an interactive chatbot with a neural network model using TensorFlow and Keras to generate relevant responses via a Flask API. – Implemented a video translation system using yt-dlp, Whisper, Google Translator, gTTS, and moviepy.

– Built a dataset from GIT history and JIRA tickets. – Developed a predictive model to estimate the impact of commits using machine learning algorithms (Random Forest, SVM, Logistic Regression, Gradient Boosting, Neural Networks), evaluated by accuracy and AUC. – Set up continuous integration with GitHub Actions, containerized the model with Docker, and pushed the image to Docker Hub.

– Implemented a Python algorithm for human detection with detection rectangle delineation and confidence. – Analyzed and implemented tools and methods aimed at optimizing and deploying the YOLO deep learning model on machines with limited computing resources.

– Modeled the energy acquisition system using Matlab Simulink. – Electronically designed the energy data box, including the ESP32 microcontroller, using ISIS Proteus 8.