Manraj Singh

Developed AI transformation pipelines using Stable Diffusion, ControlNet, and Replicate for clients including Fanta, AWE XR, and Porzellanikon; trained LoRA models to emulate branded artistic styles while preserving facial identity. Built AI workflows in Buildship for Spatial Agent with real-time object recognition (Grounded-SAM), prompt engineering, test cases, and interactive AI agents; fine-tuned models in the OpenAI Playground and prototyped AI pipelines for Marvel x Coca-Cola and Marvel superhero avatar transformations. Designed AI-powered pre-sales visuals for H&M and other brands; created an influencer pipeline for German Tourism; developed a ComfyUI-based Live Portrait video pipeline; conducted research, testing, and optimization of LoRA, ControlNet, and upscaling workflows.

Designed wireframes, process flows, and storyboards; developed high-fidelity UI mockups and prototypes; conducted user research to validate interactions and optimize AI-assisted workflows; refined UI/UX strategies based on research insights and explored visual input methods.

Developed VR environments in Unity3D with 6DOF motion, spatial audio, and gamification for immersive simulations; conducted subjective testing and user evaluations; analyzed VR interaction and optimized game mechanics to enhance engagement.