Jerardh J.

Jerardh J. AI & Computer Vision Engineer | Applied Machine Learning | Building Agentic Workflows & LLM Applications | PyTorch, YOLO, VLMs, Python | Industrial AI & Anomaly Detection | Open to Full-Time & Freelance Consulting Stuttgart, Baden-Württemberg, Germany (DE) 500 connections • 2,082 followers About I am an Applied AI Engineer and Computer Vision Specialist focused on transitioning state-of-the-art machine learning research into production-ready, scalable systems. With a robust background in industrial AI and deep learning, I specialize in building end-to-end ML pipelines—from automated data collection and annotation to model training, rigorous evaluation, and real-time deployment. My technical journey bridges the gap between traditional Computer Vision and modern Generative AI. Currently, I am designing and building Agentic AI workflows, specifically developing an AI-driven CV optimization application. This involves orchestrating Large Language Models (LLMs), Vision Language Models (VLMs), and advanced prompt engineering techniques to automate complex, context-aware tasks. Previously, as a Researcher at Fraunhofer IOSB, I drove innovations in multispectral and hyperspectral perception for intelligent sorting systems. My Master’s thesis advanced anomaly segmentation in industrial environments, where I developed novel datasets and conducted comparative analyses of state-of-the-art architectures, including unsupervised Variational Autoencoders (VAEs), deep pixel-wise one-class classifiers, Patch-Core, and optimized Draem models. Beyond just training models in a notebook, I am passionate about MLOps, stable infrastructure, and deploying autonomous systems that deliver clear, measurable metrics in the real world. I am a published researcher and an active contributor to open datasets for the AI community. 🛠️ CORE COMPETENCIES & TECH STACK: ▪️ AI & Machine Learning: Applied ML, Computer Vision (CV), Generative AI, Large Language Models (LLMs), Vision Lang