Real estate virtual tour creation from property inspection photos and survey data

Introduction Estate agents waste roughly three days per week on listing creation.

A property arrives on the market; the agent photographs it, uploads those photos to editing software, writes descriptions, creates floor plan overlays, and finally publishes everything across multiple portals. Meanwhile, potential buyers wait, and the agent misses opportunities to show other properties or qualify new leads. The real bottleneck is the middle section: photo editing, enhancement, and video creation. Most agents outsource this or do it themselves, either way consuming time that should go toward client relationships. Even with templates and batch processing, the work remains manual and repetitive. What if that entire process ran automatically? You receive raw inspection photos and survey data; a workflow captures those files, enhances and edits the images, generates a video walkthrough with a virtual presenter, and publishes the complete listing. The agent logs in to review and approve, not to create from scratch.

The Automated Workflow This workflow uses n8n as the orchestration platform because it handles file operations, conditional logic, and API calls without requiring code.

You could use Make or Zapier, but n8n's file handling is cleaner for image and video assets.

Step 1: Trigger and File Collection

Create a webhook in n8n that listens for incoming property data. Estate agents upload inspection photos to a shared folder (Google Drive, Dropbox, or OneDrive). The webhook fires when a new folder is created with the property reference number.

CopyPOST https://your-n8n-instance.com/webhook/property-listing
Content-Type: application/json { "property_id": "PROP_2026_00451", "agent_name": "Sarah Mitchell", "property_type": "detached", "bedrooms": 4, "folder_url": "https://drive.google.com/drive/folders/1ABC2XYZ3", "survey_data": { "square_feet": 2400, "year_built": 1998, "features": ["garden", "garage", "conservatory"] }
}

Within n8n, add a Google Drive node to list all files in that folder. Filter for image files only (.jpg, .png).

Step 2: Image Enhancement with Pixelcut AI

For each image, send it to Pixelcut AI to enhance quality, remove minor blemishes, and correct lighting. Pixelcut returns a cleaned-up version suitable for publication. Use Pixelcut's API endpoint:

CopyPOST https://api.pixelcut.ai/v1/enhance
Authorization: Bearer YOUR_PIXELCUT_API_KEY
Content-Type: multipart/form-data { "image": <binary file>, "enhance_type": "property", "remove_background": false, "lighting_correction": true
}

Loop this step across all images from the folder. Store the enhanced images in a temporary directory within n8n.

Step 3: Background Replacement (Optional High-Value Rooms)

Some agents want the best rooms to have neutral, premium backgrounds. Use Pixelcut's background replacement feature on 2-3 key images (lounge, main bedroom, kitchen). This step is optional but increases listing appeal.

CopyPOST https://api.pixelcut.ai/v1/background-replace
Authorization: Bearer YOUR_PIXELCUT_API_KEY
Content-Type: multipart/form-data { "image": <binary file>, "background_type": "neutral_interior"
}

Step 4: Generate Listing Description and Script

Use Claude Opus 4.6 via the Anthropic API to write a property description and a short video script based on the survey data and property type. Claude processes the structured data and generates marketing-quality text.

CopyPOST https://api.anthropic.com/v1/messages
Authorization: x-api-key YOUR_ANTHROPIC_API_KEY
Content-Type: application/json { "model": "claude-opus-4.6", "max_tokens": 1024, "messages": [ { "role": "user", "content": "Write a property listing description and 60-second video script for a 4-bed detached house built in 1998 with 2400 sq ft, featuring a garden, garage, and conservatory. Make it appealing to young families." } ]
}

Store the output in a text variable within n8n; you'll use the script in the next step.

Step 5: Create Video Walkthrough with Hour One

Send the generated script and 3-5 of the enhanced images to Hour One. Hour One generates a video with a virtual presenter narrating the walkthrough, using the images as background visuals. This step creates a professional video asset without any human recording.

CopyPOST https://api.hourone.com/v1/video/create
Authorization: Bearer YOUR_HOURONE_API_KEY
Content-Type: application/json { "script": "<generated script from Claude>", "presenter_id": "presenter_001", "images": ["image_1.jpg", "image_2.jpg", "image_3.jpg"], "video_format": "1920x1080", "language": "en-GB", "voice_style": "professional"
}

Hour One returns a video file URL. Download it and store it locally within your n8n workflow.

Step 6: Optional Meeting Notes with Cogram

If the agent conducted a site visit meeting (recorded or with a transcript), use Cogram to extract key observations and action items. This helps populate additional listing details or flags for follow-up work.

CopyPOST https://api.cogram.com/v1/meetings/process
Authorization: Bearer YOUR_COGRAM_API_KEY
Content-Type: application/json { "meeting_id": "PROP_2026_00451_visit", "transcript_url": "https://storage.example.com/meeting_transcript.txt"
}

Step 7: AI Boost for Final Image Pack

For final polish, use AI Boost to upscale any lower-resolution images to 4K and ensure consistency across the image set. This is particularly useful if survey photos vary in quality.

CopyPOST https://api.aiboost.com/v1/upscale
Authorization: Bearer YOUR_AIBOOST_API_KEY
Content-Type: multipart/form-data { "images": [<all enhanced images>], "upscale_factor": 2, "enhancement_mode": "property"
}

Step 8: Compile and Notify

Create a folder within your CMS or listing platform and upload all assets: enhanced images, upscaled images, video, description, and metadata. Send the agent a notification (via Slack or email) with a review link and approval form. n8n has built-in nodes for Slack and email. Keep the message simple:

Copyjson
{ "text": "New listing ready for review: 4 Elm Close, £425,000. Review at: https://your-cms.com/property/PROP_2026_00451. Approve or request changes."
}

Once the agent approves, a final webhook triggers to push the listing live across your property portals (Rightmove, Zoopla, etc.).

The Manual Alternative If you want more control or your property metadata is inconsistent, run the workflow partially.

Execute steps 1-4 (image enhancement and description generation) automatically, then have the agent review the description and script before proceeding to video creation. This adds one approval step but ensures quality, especially for premium or unusual properties. You could also use Claude Code (via Claude's web interface) to manually review and refine scripts before sending to Hour One, which is useful while you're testing the workflow.

Pro Tips Rate Limits and Batching. Hour One and Pixelcut both have concurrent request limits.

If you're processing multiple properties simultaneously, queue requests in n8n using the "Wait" node to space them out. Stagger image enhancement by 2-3 seconds per image to avoid throttling. Cost Optimisation. Process only 3-5 key images per property through Pixelcut, not every single photo. Choose the lounge, master bedroom, kitchen, and exterior. This cuts costs without sacrificing listing quality. Upscaling with AI Boost is expensive; use it only on your final image set, not drafts. Error Handling. Add conditional logic in n8n to handle failed API calls. If Pixelcut fails on an image, log it and continue with the next one; don't stop the entire workflow. Similarly, if Hour One's video generation takes longer than expected (it can take 10-15 minutes), use the "Wait" node to poll status rather than timing out. Approval Workflow. Always include a human approval step before going live. Have the agent sign off on the video and description; this catches mistakes and maintains brand consistency. Privacy and Compliance. If properties contain faces in photos (previous occupants, agents), consider adding a face-blurring step via Pixelcut before publishing. This is especially important for GDPR compliance in the UK.

Cost Breakdown