Thanarit K.

I build custom ComfyUI workflows and train LoRA models for AI image generation. Need product photography that looks studio-shot but costs a fraction? A consistent character across dozens of images? A style-matched brand aesthetic applied to any scene? I build the ComfyUI pipelines that make this happen. What I deliver: - Custom ComfyUI workflows: text-to-image, style transfer, image fusion, inpainting, multi-step pipelines - LoRA model training: train on your brand, product, face, or art style for consistent generation - Cloud deployment: production-ready ComfyUI setups on RunPod with 60-75% cost reduction vs standard GPU pricing - Flux, SDXL, and Illustrious model workflows Background: - 2 years of deep learning engineering experience with PyTorch, computer vision, and diffusion models related tools like ComfyUI, Kohya_ss, etc. - Designed and taught a university ComfyUI for business application workshop as invited guest instructor - Built Thai TTS voice cloning and real-time ASR systems in production - Built and deployed a serverless ComfyUI/Flux API handling production workloads I respond within hours and document everything I deliver. If you need a ComfyUI workflow built, a LoRA trained, or a generation pipeline deployed, let's talk.