Ogene Jude

Ogene Jude Machine Learning Engineer | Computer Vision (YOLO, Segmentation, OCR) | GenAI (RAG, LoRA) | PyTorch | AWS Dallas, Georgia, United States (US) 167 connections • 172 followers About I'm an AI Engineer and Computer Vision Scientist with an M.S. in Computer Science (AI & Data Science) from Kennesaw State University, Dec 2024. Currently building an end-to-end AI Highway Safety and Incident Detection Platform — training YOLOv8s on 140,000 real highway images, achieving 0.833 mAP@0.5, with a full GCP pipeline: Vertex AI for training, Pub/Sub for real-time event streaming, BigQuery for incident logging, and Cloud Run for inference. The system detects vehicles, tracks them across frames, and flags near-miss incidents in real time. My background spans computer vision pipelines, deep learning model training and deployment, OCR systems, and cloud-native ML architecture. I've built production systems at Beti LLC, won Best Data Science Capstone at KSU's DS7900 program, and organized AI/ML bootcamps mentoring students from Python fundamentals to deployment. Technical strengths: Python · PyTorch · YOLOv8/Ultralytics · OpenCV · GCP (Vertex AI, Pub/Sub, BigQuery, Cloud Run) · AWS (S3, Lambda) · FastAPI · Docker · W&B 📍 Atlanta, GA | Open to US remote 🔑 Seeking H-1B sponsorship — STEM OPT currently active 🔗 github.com/jay-ogene 🔹 Core Skills & Tools: Python, SQL, OpenCV, PyTorch, TensorFlow, Scikit-learn Generative AI (LLMs, RAG, Stable Diffusion, LangChain, Hugging Face) Computer Vision (Object Detection, Segmentation, Pose Estimation, OCR) Data Science & Analytics (Data Visualization, Predictive Modeling, Tableau) Deployment (Streamlit, FastAPI, AWS, GCP, Azure) 📌 Projects & Highlights: Built custom AI chatbots and RAG-powered apps with vector databases & LLMs. Designed Computer Vision pipelines: YOLO-based object detection, OCR with TrOCR, segmentation with SAM. Developed Generative AI solutions with Stable Diffusion, ControlNet, and LoRA. Completed Udacity Nan