Greater Heilbronn Region
I build high-performance AI systems, owning the full lifecycle from ML model development and training to research validation and scalable, production-ready deployment. Having built my foundation engineering ML pipelines for top-tier studios like Eyeline Studios (Netflix), I bring a unique, precision-driven approach to solving complex engineering challenges. My expertise lies at the intersection of deep learning architecture and MLOps. While I have deep roots in AI Perception and Computer Vision, I possess a broad, system-level understanding of diverse ML models—including Generative AI, Transformers, and VAEs. This allows me to not only rapidly prototype, develop, and fine-tune newly released SOTA architectures but also build the robust, distributed infrastructure required to serve them efficiently at scale. Core Competencies: * ML Systems & Foundational Architectures: Broad expertise in evaluating, retraining, and orchestrating diverse deep learning models to build comprehensive, end-to-end AI ecosystems. * AI Perception & Vision: Deep expertise in computer vision, 2D/3D spatial tracking, and low-latency ML model implementations. * MLOps & Scalable Infrastructure: Architecting highly scalable, global frameworks using Kubernetes, Ray, and MLflow for distributed model training and inference. * Optimization & Low-Latency Deployment: Maximizing compute utilization and conquering GPU VRAM constraints through pruning, precision conversion, and recompiling models (TensorRT, ONNX, TorchScript). * C++ Integration: Engineering custom plugins to seamlessly embed AI models directly into enterprise client applications and high-performance data pipelines.
Responsible for bringing modern AI innovations—such as computer vision and generative models—into the visual effects production pipeline. The role involves driving research and development of new AI solutions and building robust, scalable MLOps infrastructures using Kubernetes and Ray for reliable, on-demand deployment across GPU clusters. Key tasks include evaluating, refining, and optimizing AI models for production readiness, as well as designing intuitive interfaces for internal tools. Additional responsibilities cover license and compliance reviews and translating complex technical subjects into clear, accessible language for non-technical teams. The overall objective is to establish efficient, future-oriented workflows that reduce costs, streamline processes, and expand creative potential. Key Responsibilities: - Exploration, development, and integration of advanced AI technologies - Creation of scalable MLOps systems using Kubernetes and Ray - Evaluation, fine-tuning, and optimization of AI models - UI/UX design for internal AI-driven tools - License and compliance assessments - Clear communication of complex concepts to non-technical stakeholders
Integrating state-of-the-art AI technologies, including computer vision and generative AI, into the visual effects pipeline. Actively contributing to the research and development of new AI models and creating scalable MLOps workflows with Kubernetes and Ray for efficient, on-demand deployment on GPU clusters. Testing, validating, fine-tuning, and optimizing AI models to meet production standards while designing intuitive user interfaces for in-house tools. Ensuring legal compliance through license reviews and translating complex concepts into clear language for non-technical teams. Driving future-proof workflows that save time, reduce costs, and expand creative possibilities. Key Responsibilities: - R&D and integration of advanced AI models. - Scalable MLOps pipelines with Kubernetes and Ray. - AI model testing, fine-tuning, and optimization for peak performance. - UI design for in-house AI tools. - License compliance checks. - Clear communication of complex topics to non-technical stakeholders.
Research & Development of AI computer vision methods for the visual effects industry. Testing, developing and implementing latest machine learning models to increase efficiency, quality and the capacity of production pipelines. Areas of research: - Semantic, Instance and Panoptic Segmentation - Monocular Depth Estimation - Neural View Synthesis (NeRF) - Video Inpainting - Applications of various GAN Networks (CycleGAN, StyleGAN...) - Frame Interpolation, Noise Reduction and Superresolution
Avengers: Endgame Captain Marvel Doctor Sleep Dumbo Shazam The King's Man: The Beginning Ant-Man and the Wasp Sense8 Jim Button and Luke the engine driver
Doctor Strange A Cure for Wellness Renegades - Mission of Honor A Stork's Journey