Answer

Short answer:

Yes. You can use photos of construction sites or unfinished structures as a base input for AI generation. By applying an various workflows using either Chat, Create or Edit features, Rendair AI overlays finished materials, lighting, and landscaping onto the existing structural geometry, allowing you to visualize the completed project without modeling it from scratch.

1/ The gap between site photos and client vision

Construction sites are messy. They are full of scaffolding, raw concrete, and debris that make it difficult for clients to imagine the final result.

Architects often face a specific communication gap: the drawings are technical, and the site photos are unappealing. Clients struggle to bridge the two.

Rendair AI solves this by treating the site photo as a "structural sketch." The AI understands the perspective and geometry of the unfinished building but replaces the raw materials with the finished textures you describe in your prompt. This turns a progress photo into a marketing-ready visual in minutes.

> “We know it is not final, but the client needs to see where this is going.”

2/ How the workflow operates

Instead of starting with a blank canvas or a 3D model, you provide the site photo as the foundation.

1. Upload the site photo: Take a clear photo of the unfinished building. Ensure the lighting is decent and the angle captures the main volume.

2. Describe the finish: In the prompt, describe the final materials (e.g., "modern white stucco facade, timber cladding accents, glass balcony railings").

3. Edit details: If scaffolding or debris remains visible, use the Edit - Select & Modify (brush tool) to select those specific areas and prompt the AI to remove them or replace them with landscaping.

3/ What you can achieve

This workflow is not just about making things look pretty; it is about validating decisions on site.

• Facade material testing: Overlay different cladding options (brick vs. stucco) onto the actual built frame to see how they interact with real-world lighting.

• Landscaping visualization: Turn a muddy construction lot into a finished garden or plaza to show context.

• Renovation previews: Take a photo of an existing shell or ruin and visualize a complete restoration.

• Scaffolding removal: Digitally clear the site to generate "future" marketing images before the equipment is actually removed.

4/ When to use this approach

Client progress updates

Send a weekly update that shows not just what was built, but what that section will look like next month. This reduces anxiety and keeps excitement high.

Real estate pre-sales

If you are selling units in a building that is only partially up, a render based on the actual view and structure feels more authentic than a purely synthetic 3D model.

Design verification

Sometimes a built detail feels different than it did in the model. Quickly rendering over the site photo helps you decide if you need to change a material specification before it is installed.

5/ Limitations and best practices

While powerful, this workflow has constraints you should manage.

• Geometry is fixed: The AI generally respects the geometry in the photo. If you need to add a new wing or floor that does not exist yet, you may need to roughly block it out in a sketch or 3D massing before rendering.

• Debris confusion: Heavy scaffolding can sometimes confuse the AI, interpreting poles as window frames. Use the Remove Object tool to clean up these areas first for better results.

• Accuracy vs. Vibe: These renders are visualizations, not construction documents. They are perfect for selling the feeling of the finished space, but they may not perfectly align with every millimeter of the CAD drawings.