Vizard Agent

How to turn a venue's floor plans into a walkthrough video

Last updated 2026-09-01 · 8 min read

Upload the floor plans and name the venue. Vizard Agent animates the plans into a narrated route through the event, then finds the venue's own published photographs so each room appears as a real picture alongside its position on the plan rather than as a labelled rectangle.

What is the short version?

Attendees do not read floor plans. They arrive, look around for registration, find it eventually, and then spend the whole first session quietly lost. A ninety-second video that walks the route through once solves a problem no printed map in the delegate pack has ever solved.

  1. Go to Vizard Agent and upload the floor plans.
  2. Name the venue so the real photographs can be found.
  3. Say the route people will actually take.

What do you need before you start?

The plans and the venue's real name. Vizard Agent reads the room labels straight off the plan images, and naming the venue lets it go and find the hotel's or conference centre's own published photography of those same rooms.

What do you type into Vizard Agent?

Say who it is for and what they need to understand. Vizard Agent works out the sequence from the plans themselves, so the useful instruction is about the journey rather than about which parts of the drawing to show.

Prompt

Make a short walkthrough video of [venue name] so people coming to our conference can picture the layout. These two images are the [two floors] we are using. Keep it short — a whistle-stop visual of the layout. Walk people through [registration, the main ballroom, the breakout rooms and the catering foyer] in the order they will use them.

Variants worth knowing:

What does Vizard Agent actually do?

Here is the order Vizard Agent worked in on a real conference walkthrough built from two floor-plan images. The second half of the session is the part that lifts it: it goes and finds the venue's own photography and combines it with the schematics.

  1. Examines both floor-plan images and reads their room labels.
  2. Downloads the plans and checks the available tools.
  3. Generates the narration and a music bed for the walkthrough.
  4. Gets word-level timestamps for the narration so the scenes can be timed to it.
  5. Tests crops and coordinates on the plans to isolate each area.
  6. Checks the installed fonts and generates styled animated captions.
  7. Renders high-resolution frames for all nine scenes and reviews them as a contact sheet.
  8. Re-renders with generous bottom margins so the captions have clearance.
  9. Measures the loudness of the narration and the music and balances them.
  10. Generates animated clips with smooth motion across the plans.
  11. Renders the walkthrough, extracts quality-check frames and reviews them.
  12. Searches for real footage and photographs of the venue.
  13. Fetches the official gallery photos of the meeting spaces from the venue's own pages.
  14. Extracts the event-space photographs from the venue's public listings.
  15. Renders hybrid scenes combining the real photography with the floor-plan schematics.

Step fifteen is what makes it useful rather than decorative. A plan tells you where a room is; a photograph tells you what to look for when you get there, and Vizard Agent puts both on screen at once.

What does the result look like?

The walkthrough this page is written from was built as nine scenes from two floor plans, with narration, animated captions, a music bed balanced against the speech, and hybrid scenes pairing each area's schematic with the venue's own photograph of it.

The motion is a slow move across the plan rather than a fake camera flight, which is the right choice: attendees are trying to build a mental map, and a swooping camera makes that harder rather than easier.

When does this not work well?

The video can only ever show what the plans and the publicly available photographs actually contain, and venue photography is almost always of empty ballrooms staged for a brochure rather than of the corridors and stairwells where people genuinely get lost on the day.

How do you fix a result that came back wrong?

Name the room, or name the step in the route that is wrong. Vizard Agent keeps every plan crop, the narration word timings and all the downloaded venue photography, so a reordered route or a corrected room name re-renders only the scenes that depend on it.

How does Vizard Agent compare to doing it yourself?

By hand this is a designer animating a PDF, and it usually ends as a PDF with arrows on it. Vizard Agent times the moves to a recorded narration and pairs each area with a real photograph of it, which is the part nobody has time to do.

By hand Vizard Agent
The plans Shown flat Animated and timed to narration
Room photos Rarely included Found from the venue's own pages
Captions Static labels Timed to the words
Sound Music, or none Narration and music balanced by measurement

Common questions

What do I need to upload? The floor plans, one image per floor. Vizard Agent reads the room labels off them.

Why name the venue? So it can find the real photographs. Vizard Agent pairs each area with the venue's own imagery.

How long should it be? Ninety seconds at most. Vizard Agent keeps it to a whistle-stop tour.

Can it add narration? Yes. Vizard Agent records it and times every scene change to the words.

What if the plan is low resolution? It will say so. Vizard Agent needs the room labels to be readable.

Can it show more than two floors? Yes. Vizard Agent handles as many plans as you upload.

Will it invent rooms? No. Vizard Agent works from the labels on your plans and reports what it cannot read.

Can I use my own photos instead? Yes, and they are better. Vizard Agent uses supplied photography before searching for any.

Does it work for a building rather than an event? Yes. Vizard Agent builds the same walkthrough for an office, a school or a hospital wing.

Can it be captioned for silent viewing? Yes. Vizard Agent burns in captions so it works with the sound off.

Why not just send the PDF? Because nobody opens it, and if they do they cannot picture the rooms. A short video is watched once on the way there, which is exactly when the information is needed.

Can I update it if rooms change? Yes. Vizard Agent re-renders the affected scenes from the same plans.

Does it check the finished video? Yes. Vizard Agent extracts quality-check frames, reviews the layout of every scene, and measures the loudness balance between the narration and the music before delivering it.

Can it mark the fire exits? Yes, if they are labelled on the plan. Vizard Agent reads what the drawing shows and will not add routes that are not on it.