Difference Between Go To and Execute Node

Explains the difference between go to and execute flow nodes in Cognigy.ai.

Cognigy Go To and Execute Flow in UI.png

Go To and Execute Flow Node

Go To vs. Execute Flow Node

On first glance, both nodes seemingly fulfill the same purpose of switching to a different flow from the one you are currently on, but there is a small detail that makes a big difference in how both these nodes are utilized.

In this brief article, we will be going over how both nodes compare to each other and when which one is the best to use.

Go To Node

The Go To Node is used to switch over to a different flow than the one currently being run but does not return. This will happen once the node is triggered. One can select between a variable entry node or just run the target flow from the beginning. Other advanced options include reparsing intents and slots in the target flow, injecting data/text or to absorb the default context of the target flow. These functions are explained below. 

 

Cognigy Go To Node View.png

Advanced Settings for the Go To Node

Execute Flow Node

The Execute Flow Node is used to jump directly to another flow. The major difference to the Go To Node is, that when finishing the execution of the target flow, the flow will switch back to the original one in which the Execute Flow Node was triggered. From there it will continue flow execution as normal.

CAUTION

If a Go To Node is in the Target Flow (even if it's only used to jump from one point to another in the Flow itself) then Execute Flow will be stopped.

Similar to the Go To Node, you can set a custom entry point or just run the target flow from the start. As with the Go To Node you are given the option to reparse the intents and slots.

 

Cognigy Execute Flow Node View.png

Advanced Settings Execute Flow Node

Explanation of the Advanced Settings

Here is a short explanation of the advanced settings in both the Go To and Execute Flow Nodes

  • Optional Injected Text - Overwrites the user input in the target flow
  • Optional Injected Data - Overwrites the user input data objects in the target flow
  • Parse Intents - Reruns the NLU and overwrites the NLU recognition with the NLU recognition in the target flow
  • Parse Slots - The same as Parse Intents but with the slot matching
  • Absorb Context - If you have default context set in your target flow this will add it to the context of your session. 

When should we be using which node?

Simply put in one sentence: You will likely want to be using the Go To Node if the process is finished in the current flow and you it want to continue with a new goal in the conversation, the Execute Flow Node would find its best use when you want to add additional Information in the current flow but the process for this is in a different flow.

A great example for the utilization of the Execute Flow Node would be, when you have an ID&V flow you want to implement to verify the user, but after that you would like the user to continue with the normal conversation. Here you would trigger an ID&V flow with the Execute Flow Node, gather the login data or other information, and once complete return to the parent flow to continue from there.

An example for the Go To Node would be, if you want to trigger a flow that has more than one function or has a broader range of things that it can do. For example, an FAQ flow where the user might want to ask multiple questions after another, and you don’t want him to be kicked out of that flow until the user is done asking questions. Then you would have the Go To Node trigger the FAQ flow, answer the questions of the user and then have a Go To Node leading back to your main flow. Of course, it would also be possible to use the Execute Flow Node in this example, but then you have the issue that when your FAQ flow is done running, the user gets sent back to the main flow even though they might not be done yet.

Cognigy Go to In UI.pngCognigy Execute Flow in UI.png

Go To Node: Doesn’t Continue Original Flow
 Execute Flow Node: Does Continue Original Flow


Comments

0 comments

Article is closed for comments.

Was this article helpful?
0 out of 0 found this helpful