Why we built BucketDesk differently
BucketDesk is not another place to store your files. It is a better way to work with the files you already own. This page explains the problem behind that idea, the principles we design by, and the trade-offs we made on purpose.
Every Ultra feature, free for one month. No card needed to start.
The customer problem
If your business already stores files in Amazon S3, why move them into another platform just to make them easier to open?
Contracts, scans, media, exports and vendor archives land in S3 because it is durable, cheap and already part of your AWS account.
The AWS Console is built for administrators. Handing out console logins or access keys to open a PDF is a security problem, not a workflow.
Files get copied into a managed file platform, then drift out of sync with the systems that keep writing to S3. Now there are two sources of truth, two bills and two sets of permissions.
A different starting point
Traditional file platforms such as Dropbox and SharePoint provide managed environments for storing, organizing and collaborating on documents, and they do it well. BucketDesk starts from the other end: the files are already stored, so the job is making them easy to work with.
| Question | A managed file platform | BucketDesk |
|---|---|---|
| Where files live | In storage the vendor runs | In your S3 buckets, in your AWS account |
| Getting started | Upload or migrate everything first | Connect a bucket; nothing is moved |
| Who controls access | Permissions set by the vendor | An IAM role you create, can read, and can delete |
| Storage bill | Priced by the vendor, per seat or per terabyte | Stays on your own AWS bill |
| Systems that write files | Need a sync or export into the platform | Keep writing to S3; the workspace sees it |
| If you leave | Export your files back out | Delete one CloudFormation stack; the files never left |
Design principles
These decide what we build and how. When a feature cannot meet one, we say so on the page that describes it.
Bring the workspace to the data. Existing Amazon S3 storage is the starting point. Teams work with files where they already live, with no migration and no second storage destination.
Make pricing visible. Every plan price is published. Before an archive extraction or a bulk conversion runs, you see its estimated AWS cost, and the cost center shows what BucketDesk activity costs you next to what AWS bills. That work, and your storage, stay on your own AWS bill.
Make S3 approachable without hiding it. Familiar ways to browse, preview, organize and share, so nobody needs to learn AWS. The bucket, folder and access model stay visible when they matter.
Keep data close. Prefer AWS-native operations where practical and avoid moving or processing file contents without a reason. Every workflow is designed with its data path in mind.
Explain data handling precisely. Say exactly when and why file contents pass through BucketDesk, for previews or anything else. Describe the real workflow instead of claiming data is never processed.
Extend the product without breaking the promise. New capabilities such as AI, automation and sharing respect customer control, data locality and minimal movement wherever practical, and each one explains its own data handling.
For business teams
BucketDesk turns your existing S3 storage into a workspace where your team can browse, preview, organize and share files without needing to understand AWS. Your files remain in your AWS account, under the control of your organization.
No storage migration. No additional storage destination. Just a simpler way to work with the files you have.
See what teams can doFor technical buyers
BucketDesk connects through an IAM role you create from a CloudFormation template, with your own external ID and 15-minute credentials. The role reaches the bucket and folder you approve, reads by default, and gains each extra capability only when you switch it on.
Moving your files into BucketDesk-controlled storage is never a prerequisite for using the product.
Every IAM action, by planCustomer-controlled storage by design
The goal is not to claim BucketDesk never processes your data. It is to make sure that handing your files to BucketDesk-controlled storage is never the price of using it.
Data handling, precisely
Each workflow has its own data path. This is the same list we publish with the IAM permissions, so you can check one against the other.
BucketDesk signs a short-lived S3 link and your browser fetches the file directly from your bucket.
Your browser sends the file to your bucket in parts, each over its own signed link.
Streamed through the BucketDesk servers in memory for the length of the request, then discarded.
The BucketDesk document server fetches the file with a short-lived link and holds it while the document is open. Edits are saved back to your bucket.
The file passes through the BucketDesk servers to Amazon Bedrock in your own AWS account. Chats you keep are stored encrypted for 30 days after their last use.
Runs on AWS Batch in your account, reading and writing your bucket there. BucketDesk receives a summary of each job.
Deliberate trade-offs
Keeping your files where they are rules some things out and makes others harder. These are the choices we made on purpose.
Extending the product
AI, automation and sharing are where file products usually start copying data around. Here is how each one works in BucketDesk.
Questions about a document go to Amazon Bedrock in your own AWS account, through a separate role that holds only Bedrock permissions.
Account-wide storage totals come from S3 Storage Lens in your account. BucketDesk receives bucket names, sizes, object counts and storage-class totals.
Small helpers inside your account hash file names before counting reads, and BucketDesk receives the numbers.
Links expire, can be revoked, ask the recipient to sign in with an emailed link and record every open. Live links stop working when the connection is removed.
Connect a bucket with one CloudFormation stack, and take access back by deleting it. Advising clients on AWS? The partner program is built for consultancies and MSPs.