Looking for a job

Nesrine ZARRAD

@nesrinezarrad

AI researcher at University of Bologna studying generative models for medical image synthesis across datasets and metrics.

Italy

Message

At the University of Bologna’s Department of Computer Science and Engineering, I compare GAN, diffusion, and flow-based models for medical image synthesis. I evaluate their performance across datasets and quantitative metrics, and investigate how dataset characteristics affect model quality, robustness, and generalization.

At Vision Age VFX, I built a video segmentation pipeline combining YOLOv11 and SAM2 that reduced rotoscoping time by 20%. I also developed real-time mask refinement and a 4K interactive interface with under 100 ms latency.

Experience

Work history, roles, and key accomplishments

Current

AI Research

Current

University of Bologna

Mar 2026 - Present (6 months)

Compared GAN, diffusion, and flow-based generative models for medical image synthesis, evaluating performance across multiple datasets and quantitative metrics. Investigated the influence of dataset characteristics on model quality, robustness, and generalization as part of a medical imaging research publication.

Gan Diffusion Models Flow Based Model Image Synthesis Quantitative Evaluation Research

Data Analyst

Dnext Intelligence SA

Jan 2026 - Jun 2026 (5 months)

Collected, scraped, extracted, and processed 10,000+ agricultural market records from multiple data sources, building automated data pipelines and interactive analytics dashboards for forecasting and decision support. Developed the COP (Cost of Production) acreage forecasting model for Corn, Soybeans, and Wheat using feature engineering, ElasticNet regression, and walk-forward backtesting, improvi

Data Collection Web Scraping Data Pipelines Analytics Dashboards Feature Engineering Elastic Net Walk Forward Backtesting Forecasting

Computer Vision Internship

Vision Age VFX

Jun 2025 - Aug 2025 (2 months)

Built an end-to-end video segmentation pipeline (YOLOv11 + SAM2), reducing rotoscoping time by 20% in production. Developed real-time mask refinement and 4K interactive interface (<100 ms latency), enabling accurate segmentation under occlusion and fast motion.

YOLOv11 SAM 2 Video Segmentation Real time processing 4K Interface Mask Refinement

Year-End Project Internship

CLEDISS

Feb 2025 - Jun 2025 (4 months)

Constructed a full-stack Flask–React web application for automated data extraction from invoices and purchase orders, integrating OCR (Tesseract) for logo detection and Google Gemini for structured field parsing achieving 70% automation across 200+ multi-vendor documents. Orchestrated real-time verification workflows to cross-check company names and stock levels against a MySQL database, identifyi

Flask React OCR Tesseract Gemini MySQL Data Extraction Inventory Management Dashboarding CRUD

Machine Learning Internship

DIGIX

Jun 2024 - Aug 2024 (2 months)

Built a RAG-based medical chatbot using Meditron-70B (Ollama), delivering context-aware recommendations and improving response relevance by 35%. Integrated into DOCARE platform with long-term memory modules, enabling personalized insights and enhancing dialogue coherence at scale.

RAG Meditron 70B Ollama Chatbot Long Term Memory DOCARE

AI Intern

Sartex Group

Jun 2023 - Aug 2023 (2 months)

Extracted and processed data from Excel and PDF documents using Python libraries such as Pandas and PyPDF2. Utilized OpenCV for image preprocessing and analysis to support automated data extraction workflows.

Python Pandas PyPDF2 OpenCV Data Extraction Image Preprocessing