03.08.2026 aktualisiert
Premiumkunde
100 % verfügbarSoftware Engineer / Digital Twin / AI
Rosenberg, Deutschland
Weltweit
M. Sc. Computer ScienceÜber mich
Software engineer and consultant with ~10 years in generative AI and cloud. I don't just build AI solutions myself – at the Schwarz Group I introduced new technologies into business units. Technical depth meets a feel for where AI adds real business value.
Skills
AI AgentsUnity3D3D-Computergrafik3D-VisualisierungKünstliche IntelligenzAugmented RealityPythonThree.JsDigital TwinReact.jsVirtual RealityGenerative AIReact NativeDocker
I'm a software engineer and consultant with around 10 years of experience at the intersection of generative AI, cloud, and enterprise processes, fluent in German and English.
Technical skills & tools. My core language is Python (plus C# and HTML/CSS/JS, including React, React Three Fiber / R3F and WordPress); with React I've already built my own apps that integrate both local and remote AI. In AI/ML I work with PyTorch, TensorFlow and ComfyUI, covering generative text-/image-to-3D, AI classification (object detection + NLP), and local, privacy-first on-device AI. For 3D/XR I use Blender, Autodesk VRED and Unity, with hands-on experience in photogrammetry, Gaussian Splatting, and automated 3D data optimization. On the cloud and MLOps side I build with Docker, Docker Compose, Kubernetes, FastAPI, Celery, AWS, RunPod and CoreWeave (Argo Workflows), including full CI/CD with automated tests and deployments and GPU-enabled containers. I'm a certified Scrum Master and work day-to-day with Jira.
Professional experience. At rooom AG (since 2022) I design and build production-grade GenAI and cloud services – a serverless generator that creates 3D models from text/image prompts, AI classification pipelines, serverless render pipelines, and multiple 3D data-processing services (photogrammetry, lightmaps, automated 3D data optimization, format conversion), plus the surrounding CI/CD and container infrastructure. In 2026 I delivered a smartphone-based 3D vehicle-scanning solution using Gaussian Splatting (Carvatar) and a local, privacy-first AI image editor.
Earlier, as a Business Consultant at Schwarz IT (2019–2022), I introduced new technologies into the business units of a large corporate group (Lidl, Kaufland) – identifying and evaluating use cases, running requirements analysis, delivering web-based solutions, and managing vendors and stakeholders. Before that, at csi GmbH (2016–2019), I worked as a software engineer and project manager, building review, inspection and collaboration applications for automotive clients including Daimler, Volkswagen, Porsche and Audi.
I hold an M.Sc. from the University of Würzburg (2016).
Sprachen
DeutschMutterspracheEnglischverhandlungssicher
Projekthistorie
As a Software Engineer at rooom AG (since 2022), I design and build production-grade generative AI and cloud services that turn 3D data into scalable products. My work spans the full stack, from research and prototyping to deployment and the surrounding infrastructure.
Key contributions:
- Built a serverless GenAI service that generates 3D models from text and image prompts.
- Developed AI classification of 3D models (object detection + NLP), including benchmarking of different neural networks, models and rendering approaches.
- Designed a serverless cloud render pipeline (Blender, GPU render farms) for photorealistic product images, cost-optimized as only compute time is billed.
- Delivered multiple 3D data-processing services: photogrammetry (with job dashboard and QR-code calibration), lightmap generation, automated 3D data optimization and format conversion (e.g. glb to usdz).
- Created numerous Blender add-ons connecting these cloud services directly from Blender (upload, optimization, lightmaps, scene import/export).
- Set up the CI/CD and container infrastructure: GPU-enabled Docker containers with Blender (bpy), FastAPI/Celery services with web dashboards, automated tests and deployments on git push, running on AWS, RunPod and CoreWeave (Argo Workflows).
Tech: Python, PyTorch, TensorFlow, ComfyUI, Blender, Docker, Kubernetes, FastAPI, Celery, React.
As a Software Engineer at Carvatar (2026), I worked on a solution that lets users scan vehicles with a regular smartphone using Gaussian Splatting and share the resulting 3D captures on a platform.
Key contributions:
- Optimized the existing capture and reconstruction pipeline and introduced additional processing steps to increase the quality and performance of the resulting Gaussian Splatting captures.
- Worked on the internal dashboard used to manage and monitor scans and processing.
- Contributed to the app that lets users capture, upload and share their scanned vehicles.
- Manually cleaned and refined splats to remove artifacts and improve the final visual result.
- Focused on making a cutting-edge capture technique accessible with consumer hardware, without specialized rigs or scanning equipment.
Tech: Gaussian Splatting, Python, 3D reconstruction and optimization pipelines, web 3D (React / React Three Fiber), cloud processing.
As a Software Engineer for the Schwarz Group (Lidl / Kaufland, 2025-2026), I implemented web-based 3D experiences and advised on the infrastructure needed to run them reliably at scale across mobile and desktop.
Key contributions:
- Built interactive web 3D experiences for mobile and desktop, with a focus on performance, load times and a smooth user experience across devices.
- Optimized 3D scenes using the latest technologies and techniques, reducing asset sizes and improving rendering performance without sacrificing visual quality.
- Consulted on the deployment infrastructure for web 3D spaces, from asset pipelines and hosting to delivery, so that 3D content could be published and maintained efficiently.
- Worked closely with stakeholders to translate business requirements into technical solutions and to bring modern 3D capabilities into a large retail group.
Tech: React, React Three Fiber (R3F), JavaScript, WebGL, Blender, Gaussian Splatting, glTF/GLB, 3D optimization pipelines.