Foldr does this: connect your SMB file servers, SharePoint and OneDrive (and Teams, Google Drive, Box and Dropbox if you use them), and one search box looks through all of them, by file name and by what is written inside the files, with OCR for scans. Who sees which result depends on a setting you choose for each location, and by default a result is shown to anyone who can use that location in Foldr, so the second half of this page is about that setting.
Foldr uses two routes, and which one a location takes decides how much it can do.
A crawler reads the files you have told it to read and keeps what it finds in a search index that Foldr holds, so a search is answered from the index and results can carry snippets. You choose how much to read for each location, from file names only, to document contents, to contents with OCR for scanned PDFs and photographs.
Foldr can crawl a SharePoint or OneDrive location into its own index, so it is searched exactly like a file server and gets OCR, custom fields and meaning-based matching. Or you can point a location at the provider’s own search, which needs no index and no crawl, though you give up the parts that depend on Foldr having read the files itself.
Either way the files stay where they are. Foldr doesn’t move them (the index holds extracted text and details about each file), and a search that crosses a file server and a SharePoint library comes back as one list. On top of keyword matching you can switch on search by meaning (it needs an embeddings provider you bring), filter by custom fields, date, type and size, and ask Grace to refine the search you have open.
This is the question to ask any search product, so here is the full answer. Each location has a setting called “Who sees search results”, with three choices.
A result appears for anyone who is allowed to use that location in Foldr, except where the location is crawled as each person (see below). That includes files they could not open on the file server itself, because the default does not look at the permissions on individual files. It suits flat locations where everyone with access can see everything.
Foldr records the permissions on every file when it crawls, so one crawl serves everybody and results are filtered per person when they search. Foldr captures per-file permissions for SMB file servers and SharePoint sites, and you have to switch the setting on. OneDrive personal drives and Teams locations capture none per file, so on those the setting filters nothing per file.
Where every person has their own area, such as home folders, the crawler can index the location from each person’s point of view, so each person’s results come from what they could open. It repeats the crawl once per person, so it suits personal areas better than a large shared library. It only helps where the connection to the storage is made as the person, so what can be read really changes from one person to the next: with a Microsoft service account it has no effect on SharePoint, or on OneDrive and Teams when the account’s subject is fixed rather than taken from each person, and in those cases Foldr crawls once.
The last choice applies no per-file filtering and no crawl-as filtering, so every hit in the location shows to anyone who can search it. A person’s access to the location and the separation between your tenant and others still apply. It is there for locations where that is what you want, and it is a deliberate choice, never a default.
Search answers from the permissions stored in the index, and those are refreshed when Foldr's crawling processes a file or folder successfully. A change is therefore not instant, and a scheduled crawl alone does not guarantee that a removed permission has reached search (a failed permission read, for example, keeps the stored one). For locations with sensitive subfolders, check the crawl activity after you change who can see what. On locations that use the provider’s native search, the provider's own search answers the query.
Anything searched through Foldr's index has to be crawled first, and the first crawl of a large location takes a while. Crawls run on a schedule you set, and the default is never, so someone has to set one. Locations that use the provider's native search need no crawl.
Scans and photographs are only searchable when OCR is on for that location. OCR adds crawl time, so switch it on where there are scans and leave it off where there are only Office documents.
On other storage (Google Drive, Box, Dropbox, S3-compatible and the rest) Foldr's index does not capture per-file permissions, so anyone who can use the location can find what is in it, unless it is crawled as each person.
On Foldr Cloud, reading file contents is part of the Business plan and up, and Essentials searches names and metadata. On a self-hosted appliance we recommend running search on a dedicated server.
The setup steps are in the knowledge base: Setting up Foldr Search, Indexing a share for search and Foldr search tips.
Try it free for 30 days, no credit card. Or get in touch about self-hosting.
