Speckle Cloud & Connectors

Speckle Help
, ,
1

Hi everybody,
I have a specific question regarding the functionality of the Speckle connectors, if we are using the cloud hosted SaaS enterprise model.

I’ve tested the workflow using a test account, but I’d like to better understand how this would work if Speckle were integrated into our platform.

Our setup is that our users would sync their models using the Speckle connectors (AutoCAD or Revit). This would require them to download and install the relevant plugin in their CAD software, using the default Speckle Cloud URI, and authenticate with their own Speckle accounts (they would register themselves).

From there, users would publish their models to projects and models within Speckle. This is where things become unclear for us: how can we align Speckle projects/models with our own CDE structure?

More specifically:

  • Is it possible to enforce that users publish to a predefined Speckle project or model?
  • Alternatively, can our CDE be synced with Speckle in a way that restricts users to selecting only from a predefined list of projects/models?

We are also considering a hybrid approach: using Speckle Cloud, but separating our users into a dedicated custom server URL (instead of app.speckle.system). The idea would be to isolate smino users from the general Speckle user base and potentially apply additional constraints or customizations to better enforce our workflow.

Ideally, the use of Speckle would be mostly transparent to our users, they wouldn’t interact with it like a typical standalone Speckle user. We want to use the full capabilities of the connectors while maintaining control and alignment with our CDE, but this currently appears to be a technical obstacle.

Any recommendation whether this is feasible, or how others approach this, would be greatly appreciated. I’m pretty sure you have solutions for this integration.

Thanks in advance.
Have a great day
Christoph

2

Within a Smino workspace on Speckle cloud (app.speckle.systems) your users would already be entirely segregated.

When publishing from a connector the list of projects are only those where that authenticated user has explicit access for viewing or editing.

If one of your users is an existing Speckle user or chose to become so, they are responsible to select
A) The target server
B) Their target Workspace
C) The project they wish to upload to

Each of these are a permissions envelope.

From our years of supporting Speckle any assumption a user is only using a single server is insufficient. Beyond that all our development assumes a secure authentication only to the resources a user has been explicitly given access. This has been the case since v1 Speckle and is true for both our public server and any self deployed server.

Control beyond that is currently out of scope.

3

Thank you jonathon for your fast and helpful reply!
Which brings me to some further questions:
We are currently exploring an integration between speckle and smino and would like to ask about user account handling and authentication.
We are interested in whether it is possible to automatically connect our users to speckle, either by linking existing speckle accounts or by programmatically creating new speckle users and associating them with smino accounts.
In addition, could you please share how the authentication workflow would look in this case? For example, how user authorization, token exchange, or account linking is handled when integrating an external system with speckle.
Any guidance or documentation you could share with us would be appreciated.

Thank you
Christoph