DanHendriksen Moderator

Check this out: https://www.youtube.com/watch?v=B5AA03yUlr4

What is your default model for image to text in the AI Service Layer settings in Admin?

That's a great way to handle it. You can also find a function in the Workflow Executions package that will give you metadata about the workflow. I know the team has "system variables" on the roadmap, where some of these things will just be automatically there as variables that can be used…stuff like who started it, when,…

In the training, one of the examples you will create involves sending an email to people. You don't need to take the training if you don't want to, but in the hundreds of customers I've worked with on Workflows it's been unanimously good feedback that it has helped shorten the learning curve. It sounds like Coleman has…

If this is something you just want to do once, you're better off just looking at the list of people you've shared the app with (look at the card or look at the asset in the asset library) and writing them an email. Workflows are best for processes that you want to perform frequently, and want to automate those processes.…

Michelle's suggestion is a good one. You could do this with a workflow, but it would be a custom function to do the alert sharing bit. You'd also need to sort out the logic and determining how an alert (they can exist on datasets as well as cards) is related to a dashboard, which alerts should be shared and which…

Yes, the user who runs the workflow is the user who performs all the tasks. If they don’t have permission to do that the task will fail. In situations like these, our support team can setup a workflow to always run as another user (IE: you). Just understand it will always show it was ran by you, and all the ramifications…

It should be 1,000, and support can make it whatever you want…you just need to reach out to them.

John, you can take guitar…I'll tickle the ivories!

I'd suggest checking out the getting started with Workflows training. That said, all you need to do is create a workflow and customize the start form to be whatever form you want the user to fill out. Then your workflow will have one step - sendEmail. You will use your variables to craft the email recipients, subject line…

Notifications can be configured on a user by user basis.

Or, you can turn email notifications off, and then at the end just send them one email as a step in your workflow that tells them to go check the queue.

You can turn the email notifications off, and just let them check the queue.

We support processing PDF's in the Image To Text end point. I haven't had a chance to test it, but have been told it went live 3 weeks ago.

In your step where you gather employee name and manager, don't you still have the association of employee ID to manager? While it's possible to aggregate all the requests for one manager and roll them up in to one task, it's going to introduce quite a bit more complexity.

A - you can change the SMTP settings in Brand Kit, but note that it affects every single email the Domo system sends. B - You cannot with Domo's built in messenger. There is an option Reply-To parameter that will all you to hit "reply all" and have it send an email to everyone who originally received it. In these cases…

The forbidden error makes it look like your user doesn't have permission to use whatever tools you told it to use. Do you have access to AI?

So you just want a function that will create a new task in a project?

Larry, it doesn't have anything to do with the workflow. If you were to edit the webform and add a row, you'd be in the same boat. The workflow uses the same API end point. The dataset is stuck at indexing. This is an adrenaline thing. You can try manually indexing the dataset via the API (easiest is probably with the…

The API call to the workflow status should return a real time status of the dataflow. The dataflows status isn't cached or stored anywhere…

The thing is, it has no idea what you're going to have in there. This isn't a limitation I think we can work on, becuase of the nature of an object it can contain anything…and there is no object for us to inspect/understand until you've passed the object in to the workflow. That's why we allow you to edit an object to…

If the dataset is stuck on indexing, it will resolve itself…that isn't related to the workflow or anything like that. It's a Domo Adrenaline issue. It should be infrequent and should resolve itself over time. Is it still hung up here? The indexing step is where Adrenaline processes the append chain and stitches everything…

To elaborate, the reason you can't select comment is because that is looking at your TEXT variables, to pass the value from that in. You don't have an appropriate variable to pass to the property object, so hit custom and just type comment in.

No - just do Custom and type in comment. And your value is correct, except you don't need to encapsulate everything in quotes. Also, use the "Insert Variable" instead of typing them out yourself that way. I wish we could shortcode them like that, but we can't right now. Use "Insert Variable" to plug it in.

Good stuff. Glad you got it up and running.

So your workflow will have a required start parameters that is an object, and that will get written to a variable. The object will get passed in the call to start the workflow. Then you'll use the "get Object from Object" function to remove that comment object. That will create a second variable that is an object called…

So the start parameter for your workflow should be an object, that looks like that JIRA payload. It will have the following children: issue key (string) comment (object) summary (string) project (string) I would probably use a delimiter other than , because , is likely to appear in the user input data. You'll need a step…

1 - That works. 2 - No, they do not. It will write them to whatever dataset you instruct it to. 3 - You could have a function make a UUID and then include it in what you pass to the dataset. You can also grab metadata from the workflow, such as who started the workflow and include it in what you write to the dataset.

Here's a screenshot if the inputs

Here's the function.