Programmatically updating answer for an unanswered question in Custom Question Answering - azure

I have created a custom question answering project using Microsoft Conginitive Services and deployed the same.
Now I would like to create an UI where an user can see the unanswered questions and update the knowledge base with appropriate answer using the UI itself.
From the deployment details, I am able to extract the Ocp-Apim-Subscription-Key. But I am little confused about the appropriate way of API call in my UI to track and update the unanswered questions.
Could not find many write-up or examples regarding this. Please guide me regarding the same.

It depends on the questions answering resource type you are using, as this service has evolved.
Azure Cognitive Service for Language (end of 2021)
This is the newest version of QnA Maker (as of late 2021), now called "Custom question answering" being part of Azure Cognitive Services for Language.
This service has a better offering in terms of API:
Service documentation is here
API documentation here
The "active learning" feature is offered through Language Studio portal (see doc).
But if you want something custom, you might be interested in the Authoring API: programmatically, you could use "Get QnAs" operation will give you all QnAs, and the RetrieveQnaRecord object included in the response contains a SuggestedQuestionsCluster object which could lead to interesting feedbacks to add.
QnA Maker (v4.0)
QnA Maker 4.0 has a REST API documented here.
In addition, you can directly check some calls using the API exposition in the right region. Example for West US here
=> For this service, there is no "feedback" solution exposed. The API allows to update a knowledge base, but you will not have an operation providing unanswered questions.

Related

Best practices to handle documentation in Azure

Cloud-Platform: Azure
Resource: Logic Apps and Functions
Hi everyone,
Kontext:
We have developed many Logic Apps and Functions in our company so far. We do not use a proper or standard methodology to handle the documentation of the user requirements and also the required technical documentation of these logic apps and functions. For technical documentation I mean all the documentation generated during the development life-cycle.
We use also Jira to assign the logic apps inside the company and also we are using X-RAY for testing. We have also Confluence for documenting our guidelines to development.
Question:
I would like to know what are the best practices to handle the documentation and very important how to keep it up-to-date?.
Is there an functionality in Azure, Jira or Confluence that we can control easily the documentation of our logic Apps and functions?
It depends on the source code repository and the deployment pipeline that you are using to deploy the Azure functions and logic apps. If you are using GitHub or Azure DevOps, then you can make use of the GitHub/DevOps wiki to create documentation about logic apps or function projects just by creating readme files. Please review the following articles which will help you with the documentation
https://docs.github.com/en/communities/documenting-your-project-with-wikis/about-wikis
https://learn.microsoft.com/en-us/azure/devops/project/wiki/wiki-create-repo?view=azure-devops&tabs=browser

Retrieve azure update management status using REST API

I'm having trouble with retrieving azure update management details programmatically. I couldn't find any REST API or CLI method to get the data from update management view. There were only powershell scripts.
These are the details I want to retrieve via REST API
As shown in below screenshot, currently I see only below highlighted REST API's that are related to update management. Here is the document reference for the same.
If these REST API's doesn't help to accomplish your requirement then if interested you may share it as feature request in this uservoice or feedback forum.
In general, Azure feature team would check feasibility of a feature request, prioritize against existing feature backlog, add in roadmap as appropriate and would announce and/or update the related Azure document once a feature request is addressed.
On the other hand, with the help of this references, you may find the same information in update management related kusto tables which you can query with the help of this REST API.
Here is the reference link for the same issue which was addressed in Microsoft Q&A

Guideline for Inter Service Communication with principal propagation with Cloud SDK on CF

We are trying to set two micro services where the principal needs to be propagated from the first microservice(Login happens here) to the second one (This uses cloud SDK to talk to S/4 Hana Cloud). Has this particular scenario been explored before / is there any existing guideline the Cloud SDK team suggests.
Anirban
The Cloud SDK capabilities related to Principal propagation in conjunction with OAuth flow are nicely explained in this blog article by Marco Dahms.
I hope you'll find an idea of how to adapt it to your use case. There are additional helpful links in the comments section of the blog.
If you give more details or what solutions you have already tried we are happy to help you further.
Cheers,
Artem

Using which Azure API my software can identify if it's running on Azure infrastructure and not elswhere

We are going to create Virtual Machine Image on Azure Marketplace. This VM Image will have our software, which has default feature and advanced feature. We don't want to ship/install software with advanced feature.
We want to write a service using Azure API, which will first check VM created using this Machine Image or software is running on Azure only and not else where before enabling the advanced feature.
I've found answers to similar questions but all were for WebService. I've also gone through API references provided on Docs section but there too I couldn't find any api solving my purpose so raising this question.
Please suggest.
Probably the easiest way would be to attempt the instance metadata service query. You can find more information here:
https://learn.microsoft.com/en-us/azure/virtual-machines/windows/instance-metadata-service

Apply the latest runtime updates to improve QnA Maker results - re-migrations required?

We received an e-mail today from Microsoft with "recommended actions" regarding the QnA Maker runtime upgrade.
Apply the latest runtime updates to improve QnA Maker results
You’re receiving this email because you currently use the generally available
Cognitive Services QnA Maker service. We’ve made some important
runtime updates that will require you to re-migrate from the free
preview version. These updates will improve the quality of results
from the QnA Maker service. Recommended action To benefit from the
updates, we recommend that you follow these steps:
Upgrade the QnA Maker runtime as documented.
If you have migrated a knowledge base from the free preview, please re-migrate by following these steps.
When the QnA Maker service became generally available, we migrated from the preview version to the GA version. Since then, we heavily edited our knowledge base(s), therefore it has far more content than the version from the preview version.
The recommended action now states that we have to re-migrate our knowledge base(s) from the preview portal, which would make our recent changes in the QnA Maker service unusable.
Does anybody know the background of these recommended actions? Or is there any way to "upgrade" the QnA Maker knowledge base(s) without re-migratring from preview version?
Thanks in advance!
It is just a runtime update, all you need to do is restart the App Service as stated in the documentation referenced in the email you received and that is it, you don't need to worry about the preview stack anymore as Microsoft is encouraging everyone to migrate to and test the new GA stack as early as possible before the preview stack is no longer sustained.

Resources