AI & Computer VisionPrototype
Smart Aquaculture AI / Firmatech
Smart aquaculture MVP combining AI fish analysis, pond monitoring, alerts, mobile workflows, and an admin dashboard.
Connects pond data to fish counting, health detection, biomass estimates, generated reports, alerts, and operator interfaces.
FastAPIPythonYOLOv8OpenCVRoboflowCentroid Tracking+12
2024-2025AI/Research/Testing and AI Module & Data Analytics contributor in the Junction Algiers hackathon team, building on earlier smart fish farming research.
AI & Computer VisionPrototype
Drawzy AI / MobAI Hackathon
Hackathon drawing game where the mobile app sends sketches to a FastAPI backend for object prediction and round scoring.
AI integration inside a product flow: canvas export, preprocessing, CNN sketch classification, RL-style scoring, mobile screens, and a team landing page.
FastAPIPythonTensorFlowKerasStable-Baselines3Gymnasium+7
2025AI contributor for the MobAI hackathon at ENSIA, focused on sketch recognition and reinforcement-learning scoring while the team built the mobile and landing surfaces.
AI & Computer VisionExperimental
PathOS / Prostate Lesion Detection
Whole-slide-image pipeline for ISUP grade prediction with tissue segmentation, patch registries, foundation-model embeddings, artifact gating, MIL aggregation, cribriform analysis, heatmaps, and tracking.
Medical AI research work covering gigapixel WSI preprocessing, GPU feature extraction, artifact-aware MIL modeling, ablation evaluation, and explainability outputs, framed as experimental decision-support research.
Python 3.12PyTorchCUDAOpenSlideOpenCVHistoEncoder+9
2025-2026Research and ML engineering work on a prostate lesion detection pipeline. Repository links stay private while the portfolio presents architecture, model results, and experiment artifacts.
AI & Computer VisionExperimental
Pneumonia Detection Lab
Educational chest X-ray lab comparing feature-engineered logistic regression and CNN models for Normal vs Pneumonia prediction.
Covers dataset preparation, preprocessing, handcrafted descriptors, notebook training, saved models, metric plots, and a desktop inference UI.
PythonJupyter NotebookTensorFlow/Kerasscikit-learnOpenCVscikit-image+7
2024-2025Academic team contributor with Chawki Belhadid, presenting the public repository as educational medical-imaging work rather than a diagnostic product.
AI & Computer VisionExperimental
Automated Coin Counting
Classical computer vision GUI that segments coin images, validates circular regions, groups coin sizes, and reports detected-vs-actual count accuracy.
Covers adaptive preprocessing, thresholding, filtering, connected components, CSV-backed evaluation, and a usable desktop workflow.
PythonTkinterPillowNumPySciPyPandas+4
2024-2025Academic computer vision lab with Chawki Belhadid, focused on classical image processing, count validation, and a Tkinter desktop interface.
AI & Computer VisionPrototype
Image to Text OCR Application
Full-stack OCR app that converts image text into editable text in the browser and stores images plus extracted text in MongoDB.
Combines Tesseract.js client-side OCR with an Express backend for upload, review, and persistence.
React.jsTailwind CSSTesseract.jsHTML5Node.jsExpress.js+3
2024Full-stack developer: built the React frontend with in-browser Tesseract.js OCR and the Node.js/Express backend for image upload and MongoDB persistence.
AI & Computer VisionPrototype
PropreModelia: Land Price Predictor
PropreModelia predicts Algerian land prices from superficie and secteur using custom linear regression models, a FastAPI API, and a Next.js interface.
Covers data collection, model training, REST API design, and a deployed prediction interface.
Next.jsTypeScriptMaterial-UITailwind CSSFastAPIPython+2
2024Full-stack developer: trained the regression models, built the FastAPI inference backend, and implemented the Next.js frontend.