The main objective of the end-of-study project is to discern the essential Machine Learning (ML) capabilities and algorithms required to process logs and traces of observability data for critical operational use cases. Then, implement and evaluate the ML Algorithms previously identified and compare them to existing tools on the market, such as Dynatrace, Elastic Suite, and similar.

Collaborated with a team to develop an application for fraud detection in medicine packages and tablets.• Performed Web Scraping to gather a dataset of images and information about medicine packages and tablets.• Extracted text from images with OCR tools and trained a custom spaCy model on the processed data to predict labels.• Used Jaccard Similarity scores to detect fraud between the information on the packaging and the tablets.• Built a dashboard to facilitate user interaction. Tools : BeautifulSoup4, Power Query, Pytesseract, EasyOCR, OpenCV, spaCy, Tensorflow, Streamlit.

Developed a medical chatbot and a Named Entity Recognition (NER) system to extract disease information from medical courses.• Extracted text from images and medical PDF files.• Trained a custom spaCy model on medical annotations to classify the extracted text.• Fine-tuned a BERT base model on a wide dataset of questions and answers between patients and doctors to create a Q-Achatbot.• Displayed the chatbot with a Streamlit application.Tools : Pytesseract, pdf2image, pdfminer, spaCy, BERT, Streamlit.

This internship's main objective is to use new artificial intelligence technologies for the early detection and automatic detection of the keratoconus disease.Link to my work : https://www.kaggle.com/code/imenbakir/keratoconus-early-stage-detection