Build Custom ComfyUI img2img Workflow (Realistic + Stylized Outputs)

I'm looking for an experienced ComfyUI user or workflow developer to create a custom img2img pipeline that transforms user photos into stylized versions with realistic detail and optional artistic styles.

? Requirements:
Works with 1 to 3 people in a photo
Supports text prompts, negative prompts
Ability to easily load LoRA models (cartoon/fantasy/cosplay styles)
Optional: ability to replace or blend in a custom background image

Three preset levels of transformation:
Minimal edit (enhancement, light stylization)
Medium edit (partial costume/background style)
Full transformation (cosplay/fantasy look)

?️ Input/Output examples:
I have 2 example images that show:

Prompt-driven transformation (e.g., into a ship captain on a yacht)

Visual-only transformation of group photos (without prompts)

? Deliverable:
A working .json workflow file compatible with ComfyUI

Easy to swap prompts, LoRA styles, or backgrounds

Optional bonus: batch or folder-based input support

? Budget: flexible based on experience and scope
Please include a timeline and examples of previous ComfyUI work if available.