Skip to content
VispullGet early access
Visual search for creative studios · Private beta

Find any image in your archive by what's actually in it.

try“the sunlit room with the tall windows from the museum proposal”

Search your whole archive by describing what you remember. Vispull reads what's inside each image and searches across Drive, Dropbox, and uploads, without copying anything out.

Private by default. Never trains on your work.

Reads your imagesSearches where your files liveNever trains on your work

By the numbers

~10 hrsMcKinsey Global Institute, The social economy: Unlocking value and productivity through social technologies, 2012.

per week

spent by knowledge workers searching for information and assets they already have.

With Vispull, that search collapses to seconds.

80%+IDC and Gartner briefings on unstructured enterprise data, 2022 to 2024.

of enterprise data

is unstructured. Most of it is images, video, and documents no filename can describe.

Vispull reads the content, not the folder tree.

0Vispull product design principle. Files are read in place via connected sources.

files moved

Your archive stays in Drive, Dropbox, and the folders where your team already works.

Vispull layers on top. Nothing is copied out.

?
Who remembers where we put that image?
...
?
Where's the rendering we shared with the client?
...
?
Where's the image from the museum proposal?
...
?
What's the filename of the room with those really tall windows?
...

Creative work isn't remembered by filenames. It's remembered by materials, conversations, projects, spaces, and moments.

Search your content the way you remember it.

Vispull · one search8 matches
The sunlit room with the really tall windows, from the museum proposal.
Here's what I found, read by what's actually in each image.

The old way, and the Vispull way.

Filenames like Final_v3_REAL_final.jpg
Search by what's in the image, in plain language
Assets scattered across Drive, Dropbox, and local disks
One search across every connected source. No migration
"Ask the person who made it" is the retrieval system
Anyone on the team finds it in seconds, without pinging the designer
Re-shoots and re-renders because the original can't be found
Recovered assets stay in circulation. No duplicate work
A richly furnished interior with distinctive materials and lightIMG_4821.jpg

Recognition

A general search sees

IMG_4821.jpg · Screen Shot 2024-03-14.png

Vispull sees

sunlit interior, oak floor, tall arched windows, warm sand palette, calm mood

  • subject
  • material
  • palette
  • composition
  • mood
  • the reference it resembles

Filenames and folders describe nothing. Vispull reads the picture.

How it works

Three steps from archive to answer.

01

Connect your sources

Point Vispull at Drive, Dropbox, or drag in a folder. Nothing leaves your workspace.

02

Search by what's in the image

A subject, a material, a color, a mood, a reference. Text, image, or voice. Vispull reads the picture, not the filename.

03

Pull the collection

Refine, save, export a share-ready board with credits and source folders intact.

It actually understands what it's looking at.

Most tools match filenames or generic tags. Vispull recognizes what's in the image, the way a person on your team would. And it learns your archive as you use it, so it gets to know your references, your style, and the way your studio talks about its own work.

All your files. Private.

Your work stays in your Drive and Dropbox. Vispull makes it searchable. Only for your team. Nothing is copied out, shared across accounts, or used to train AI models.

How we handle your data

From the founder

"I spent years inside a working studio surrounded by deep visual archives. I watched talented people lose afternoons hunting for images they knew existed, because no filename could hold what was actually in the picture. General tools still can't read a picture the way a designer does. So I built one that can."

Bruno Marsino, founder

Built for

Studios that live inside their references.

Vispull is shaped around teams whose work is visual, archival, and messy by nature.

01

Design studios

Brand, editorial, and product teams juggling moodboards across drives.

02

Architecture & interiors

Practices searching precedent by material, light, and spatial feel.

03

Large reference libraries

Studios with years of shoots, renders, and references piling up.

FAQ

We're opening workspaces in small batches during private beta. Add your email and we'll send one invite when your workspace is ready. No newsletter, no drip.

Google Drive, Dropbox, and folders you drag in. Vispull reads the images where they already live, so nothing gets re-uploaded into a separate platform.

General search matches filenames and generic tags. Vispull recognizes what's actually in the picture: subject, material, color, composition, style, and the references it resembles. You search by what you remember.

Yes, it's private, and no, it never trains on your work. Your files stay in your Drive and Dropbox, searchable only by your team. Nothing is copied out, shared across accounts, or used to train AI models.

Yes. You can export your saved boards and remove Vispull's access at any time. Because your files never move, disconnecting simply ends the ability to search them.

Your images always remain in your own Drive and Dropbox, untouched. If Vispull ever winds down, you lose nothing but the search layer itself.

Pricing is being finalized during private beta. Early workspaces get first access and founding terms. Join the list and we'll share details before you're asked to pay.

The ~10 hours/week figure comes from McKinsey Global Institute, 'The social economy: Unlocking value and productivity through social technologies' (2012). Estimates that 80%+ of enterprise data is unstructured and growing come from IDC and Gartner briefings. Tap the small info icons in the By the numbers section for the direct source of each figure.

Get the invite before your next project.

We're opening workspaces in small batches. One email when yours is ready. No newsletter, no drip.