Trying out Flows
The Try Out functionality in Teneo Studio allows to test and ensure that the Natural Language Interaction (NLI) application behave as expected. The user can test the NLI application as often as needed simulating a chat session, and it is advisable to try out every Flow to make sure that they are working properly.
Teneo Studio offers two different views of the Try Out, a simple Try Out available in the right-side of the Teneo Studio window, and an advanced Try Out available through the Try Out button in the top ribbon or the Advanced button available in the simple Try Out view.
Simple Try Out
The simple Try Out offers the user the options to quickly test an input and ensure that the input triggers the correct Flow.
Follow these steps to try out the NLI application:
- In the Home tab of Teneo Studio, ensure the simple Try Out is open in the right side of the window or click the Try Out tab to open it.
- Now, imagine an end-user is interacting with the NLI application; type a user input in the text box and hit Enter or click Send to see what answer the NLI application returns.
If the expected answer is returned for the input it means that the NLI application is working as expected.
In this view, it is possible to Add Parameter, but for information related to the processing path, engine status, etc. the user will need to open the advanced Try Out.
Clicking the Advanced... button allows to:
- Open the advanced Try Out window
- Restart the chat session
- Load the latest version of the solution documents; learn more
- Load the stable version of the solution documents; learn more
- Clear the dialogue history.
Advanced Try Out
To access the Try Out window, click the Try Out button in the top ribbon of Teneo Studio or the Advanced button in the simple Try Out view.
The top ribbon of the Try Out window offers the following options:
Close the Try Out window
Restart this button allows to start a new conversation; this is particularly useful if testing a Flow that uses global variables as Restart will start a new conversation and reset the value of all global variables to their initial values.
The Restart button also includes the below option when clicking the lower part of the button:
- Reload Engine this button allows to recreate the Engine Try Out instance. Normally, this is not needed, but (very rarely) the knowledge base used by the Try Out Engine is not informed of a change made by a user in Studio and a reload can be necessary. Pressing Alt + Restart will also re-synchronize the Try Out Engine with the solution knowledge base.
Export Engine XML choose this option to export the Engine XML.
Replay this button allows to replay an input or several by selecting one of the inputs in the Try Out dialog view.
Clear All allows to clear previous dialogues and messages from the view.
Stable Version: Latest/Stable allows to test either the latest versions of all documents in the solution or the stable versions of documents in the solution, read more here.
The chat functionality of the advanced Try Out is similar to the described above for the simple Try Out, but the advanced Try Out also offers detailed information related to the chat, right beside the Dialog area, such as information about the dialogues, session and transactions.
- Input here the user is able to find information about the machine learning classification, annotations and input processing performed on the input.
- Path the path gives information related to the path taken from input to output, such as which Flow was raised, if any listeners were fired, transitions passed, etc.
- Output here the user finds the answer given as well as information related to the output, such as emoticons or URLs.
- State this section gives information related to the current variable values and the Engine state.
- Messages the messages displays information received during the session, in-between transactions.
Icons in the Dialog view
When inputs and outputs are tested in the Try Out, icons are displayed below the answer text with basic information of how the Try Out Engine handled the input. The icons are clickable and takes the user to the specific (Flow) node when clicked.
- Trigger this icon represents the Intent Trigger that fired when the user input was sent to Engine.
- Raise Flow this icon represents the raised Flow (when entering the Flow).
- Output this icon represents an output node from which the answer text came.
- Drop Flow this icon represents the action when the Flow was dropped (when leaving the Flow).
- URL this icon is displayed when an answer contains a URL (clicking the icon will open the URL).
- Emotion icons of emotions are displayed if these are assigned to the Flow output nodes.
ML model messages
Teneo Learn, the component which trains the machine learning model of a solution directly in Teneo Studio, sends information to the Try Out about the model training status.
There are different messages displayed related to the model training, for example Model update in progress or Model update completed. The messages are meant to help the user understand the status of the machine learning model.
When the Model update completed notification is shown, the user can test the latest machine learning model in the Try Out.
The Try Out Engine will tag each user input with the annotations generated by the model, those annotations can be seen in the Input section in the Try Out window, where the Classifications, Annotations and Processing > Input Processor Results areas provides the user with information related to the processing of the input.
Read more about the Machine Learning model here.
The model will be available for the user whose save action initiated the training, other users that are currently editing the same solution will continue using the old version until its Try Out is reloaded (a restart will not trigger a model update). Again, the Model Update Completed message will be used to indicate the model is ready for its use.
Users that connect to a solution get the latest model trained for that solution.
Sometimes, a Reload now message will be displayed above the text field of either of the Try Out views; this occurs when a change in the knowledge base or solution requires a reload of the Try Out Engine, for example after performing changes to the Ordering or assigning a new lexical resource. An Engine Reload Required message is also displayed in the lower right-part of the main Teneo Studio view.
Solution messages and warnings
In both Try Out views, messages and warnings are shown to quickly draw the user's attention to these; the information is - in the simple Try Out view - displayed as a brief message which can be clicked for the user to open the Messages section of the advanced Try Out for more detailed information.
Export a Try Out session
When right-clicking in the Try Out window, the user will be presented with different export options depending on where the user clicked. Usually the options include the possibility to copy or open a summary, open the chat log or full log, and to Clear All from information from the Try Out.
In the Dialog area if right-click is activated in the field of an input/output, the user has the option to Open Chat Log either As Text or In CSV format. Selecting one will open the available chat log in either a text editor or Excel.
Note that in the top ribbon of the Try Out window, an option to Export Engine XML is available. Here the user will, among others, be able to find information related to annotations.
Also, when right-clicking in the input/output field, the user is presented with the Replay option which allows to automatically send inputs to the Try Out again to re-test them.
Parameters may be used by a frontend when the user wants to send information to Engine that is not (part of the) user input text. Such parameters can be simulated by adding them in the either Try Out view by clicking Add Parameter.
Input parameters can be fetched and processed in Global Scripts in Teneo Studio (Solution tab > Globals> Scripts). In the user feedback example shown below, input parameters are extracted from the request sent to Engine and their values are stored in Global Variables set up for this purpose.
- currentPageUrl: It obtains the current URL from which the feedback was sent
- userFeedback: It checks whether user feedback was sent
- userFeedbackComment: It obtains the user feedback comment from the frontend.
Of course, in this example,
userFeedbackComment have been declared as variables.