AI Virtual Staging Image-to-Image API

I need AI Virtual Staging Image Generation System

Task Overview
I need an AI-based virtual staging system that can take an image of an empty room (or not empty) and generate realistic, fully furnished interiors while preserving the original room structure (walls, windows, floor, ceiling). If room already has furniture we need to delete it

Technologies that seem to be good (I have tried them for some experiments):
Stable Diffusion XL (SDXL) – Base image generation model
LoRA – For photorealistic furniture textures or FLUX + ComfyUI.
Real-ESRGAN – For upscaling and improving image clarity

What I Need
Image-to-image pipeline that takes room photo and generates a fully furnished interior.
Furniture should be realistic (not look like AI-generated sketches).
Room details must remain intact (no distortions of walls, floors, windows, or doors).
Option to apply different furniture styles (Modern, Midcentury, Scandinavian, Luxury, Coastal, Farmhouse). and room types (Bedroom, Living Room, Kitchen, Home Office)

I have runpod instance to deploy the solution

references:
https://www.virtualstagingai.app/

solution which I tried to implement
https://www.reddit.com/r/StableDiffusion/comments/1azqmmi/exploring_diffusion_models_for_furniture/ (see comments from Hanumankattu)

are you able to impalement this (close to https://www.virtualstagingai.app)

I do not need entire solution. What I need is just core api which takes image + input params (like room type and style) and returns result image. What are your thoughts? Are you able yo help with this?