AI Influencer Creation and Pipeline Setup

I am looking for an experienced ComfyUI specialist to build an original, realistic AI influencer from scratch, and set up a complete cloud-based content pipeline that I can run myself. The character will mainly be used for brand-safe UGC, TikTok Shop and affiliate content on TikTok and Instagram.

What I need:

1. Character creation and LoRA training
- Design an original, photorealistic character (not based on a real person)
- Build the training dataset and train a LoRA (Flux or another model you recommend)
- The face and body must stay consistent across images and videos

2. Image generation workflow
- Text-to-image workflow with the LoRA built in
- Lifestyle and product-holding shots suitable for brand content
- Upscaler included

3. UGC video workflow
- Image-to-video workflow (e.g. WAN 2.2 or similar), 9:16 vertical format
- UGC-style clips with lip sync, so the character can speak to camera about products
- The trained LoRA must be integrated into the video workflow (not only the image workflow), so the character's face and identity stay consistent between images and video, and throughout each clip

4. Cloud setup
- Everything deployed on RunPod (or a similar cloud GPU service). I don't want ComfyUI running on my own computer.
- Simple to operate: change the prompt, upload a product image and hit Run
- Workflows should let me switch between trained LoRA characters from a dropdown, with no reconfiguration, so I can add more characters later

5. Training and handover
- A live walkthrough session (screen share) teaching me to generate images and videos
- All files delivered: LoRA (.safetensors), workflow files (.json), and prompts

Requirements:
- Proven hands-on experience with ComfyUI, LoRA training and AI video workflows
- Portfolio showing realistic, consistent characters, ideally UGC or brand-style content
- Before final delivery, provide a test set (e.g. 20 images and 3 videos in different settings and outfits) showing the character stays consistent