Real estate listing automation from inspection photos to virtual tour to social posts

A property inspector walks out of a three-bedroom house with 47 photos on their phone. The agent spends the next two days cropping, enhancing, writing descriptions, filming a tour video, and creating six Instagram posts. By the time the listing goes live, two other properties in the same street have already sold. This is not an edge case in real estate, it is the default workflow for most agents, and it costs thousands in lost commissions every year. The good news: every step of this process can be automated. A single trigger, somewhere in your system, can set off a chain reaction that transforms raw photos into a polished listing, a professional video tour, and social media content ready to post. No manual handoffs. No waiting for contractors. No excuses about being too busy. This is what we will build today: a workflow that takes inspection photos, enhances them through AI image processing, generates a virtual presenter video tour, and automatically creates social media carousels. The entire pipeline runs in the background while you are doing other work.

The Automated Workflow

We will use n8n as the orchestration engine for this workflow, because it handles file operations and multiple API calls better than Zapier or Make, and because it runs locally if you need it to. You can also use Claude Code to build custom logic if your images or data require special handling. The flow works like this: photos land in a folder or are sent via webhook, they get enhanced and upscaled, metadata is extracted and passed to a video generator, a virtual presenter video is created, carousel content is generated, and everything is marked ready for publishing.

Step 1: Trigger and Photo Ingestion

Create a webhook in n8n that listens for incoming photos. Your camera app or cloud storage can send files here.

CopyPOST /webhook/real-estate-photos { "property_id": "PROP-2024-0847", "agent_id": "AGENT-523", "photos": [ { "url": "https://storage.example.com/photo-1.jpg", "room": "living room", "timestamp": "2024-11-15T14:32:00Z" }, { "url": "https://storage.example.com/photo-2.jpg", "room": "master bedroom", "timestamp": "2024-11-15T14:35:00Z" } ], "property_details": { "address": "42 Whitmore Lane, Bristol BS8 2HH", "bedrooms": 3, "bathrooms": 2, "asking_price": "£495,000" }
}

Step 2: Image Enhancement with AI Boost

For each photo in the batch, call the AI Boost API to upscale and enhance the image. This sharpens details, improves lighting, and removes minor blemishes without looking overdone.

CopyPOST https://api.aiboost.io/v1/enhance { "image_url": "https://storage.example.com/photo-1.jpg", "enhancement_type": "real_estate", "upscale_factor": 2, "auto_adjust_lighting": true, "remove_clutter": false
}

Store the returned enhanced images in your own cloud storage. Note the URL; you will need it later.

Step 3: Extract Property Details and Generate Narration

Use Claude Opus 4.6 via API to analyse the photos and create a description suitable for a video voiceover. Feed it the room metadata you received in the webhook, plus any property details from your CRM.

CopyPOST https://api.anthropic.com/v1/messages { "model": "claude-opus-4.6", "max_tokens": 1024, "messages": [ { "role": "user", "content": "You are a real estate marketing expert. Based on the following property details and room list, write a 60-second video script for a property tour. Be warm, professional, and highlight key selling points. Property: 42 Whitmore Lane, Bristol BS8 2HH. 3 bed, 2 bath. Asking £495,000. Rooms: living room, master bedroom, kitchen, bathroom, hallway. Focus on natural light, period features, and modern updates." } ]
}

Claude will return a coherent script. Save this as a variable in n8n to pass to the next step.

Step 4: Generate Video with Hour One

Hour One turns text into video with a virtual presenter. Send the script from Step 3 along with the enhanced images from Step 2.

CopyPOST https://api.hourone.ai/v1/videos { "script": "Welcome to 42 Whitmore Lane, a stunning three-bedroom property in Bristol's sought-after Cotham area. As you walk into the light-filled living room, you'll notice original cornicing and high ceilings that frame this generous space beautifully...", "presenter": { "id": "presenter_default_female", "style": "professional" }, "background_images": [ "https://storage.example.com/photo-1-enhanced.jpg", "https://storage.example.com/photo-2-enhanced.jpg" ], "video_format": "landscape_1080p", "duration_seconds": 60
}

This returns a video URL. Save it.

Step 5: Create Social Media Content with Mirra

Mirra generates carousels and short social videos from prompts. Send it the property details and ask for Instagram-ready content.

CopyPOST https://api.mirra.ai/v1/content { "format": "carousel", "platform": "instagram", "content_brief": { "property_address": "42 Whitmore Lane, Bristol BS8 2HH", "key_features": ["3 bedrooms", "Period property", "Natural light", "Modern updates", "£495,000"], "tone": "aspirational, warm", "call_to_action": "Contact us today for a viewing" }, "primary_image_url": "https://storage.example.com/photo-1-enhanced.jpg", "secondary_images": [ "https://storage.example.com/photo-2-enhanced.jpg", "https://storage.example.com/photo-3-enhanced.jpg" ]
}

Mirra returns carousel slides ready to download or post directly.

Step 6: Store and Notify

In n8n, write all outputs to a database or spreadsheet: enhanced photos, video URL, carousel content, script used. Send an email or Slack message to the agent confirming that the listing is ready for review and publication.

CopyPOST https://slack.com/api/chat.postMessage { "channel": "AGENT-523", "text": "Your property listing for 42 Whitmore Lane is ready.", "blocks": [ { "type": "section", "text": { "type": "mrkdwn", "text": "Property: 42 Whitmore Lane, Bristol\nEnhanced photos: 5 ready\nVirtual tour video: https://example.com/tour-video.mp4\nSocial carousel: 4 slides ready\nTime to completion: 8 minutes" } } ]
}

The entire workflow, from webhook to notification, should complete in under 15 minutes. All the agent has to do is review the outputs and click publish.

The Manual Alternative

If you want more control over messaging or creative decisions, stop the automation after Step 2. The agent can then write their own descriptions, hand-pick which photos go into the tour, and customise the video script before it goes to Hour One. This takes longer but lets you maintain brand voice consistency. You still save 80% of the labour.

Pro Tips

Rate limits matter.

AI Boost and Hour One both throttle requests.

If you are processing multiple properties per day, stagger the API calls using n8n's built-in delay node. A two-second pause between photo enhancement requests prevents timeouts.

Image size and quality.

Larger photos (over 10MB) take longer to enhance and cost more. Resize incoming photos to 2048 pixels on the longest edge before sending to AI Boost. This halves processing time and cost with no visible quality loss.

Keep your scripts short.

Hour One charges per minute of video. A 60-second script costs roughly half what a 120-second script costs. Three minutes of narration is the sweet spot for property tours; viewers lose attention after that.

Test with one property first.

Run this workflow manually on a single property using the API directly. Check the outputs at each step. Once you are confident, wire it all together in n8n and test again with a second property before running it on your entire pipeline.

Store everything.

Keep copies of enhanced photos, videos, and carousel content in a database keyed by property ID. Agents often want to repurpose content months later, and having the raw outputs saved means you do not have to regenerate from scratch.

Cost Breakdown