At the time of writing this blog post, it is hard to be entirely satisfied with the existing Azure cost control solutions such as Cloudyn or the Microsoft Azure Consumption Insights Power BI app, should you envision a very granular way of analyzing costs.
Indeed, both Cloudyn and the Power BI app help to analyze costs per subscription and even per resource group to some extent but none of these solution focuses on tags, although tags are the only way you can really tie all things together, as for instance tagging whatever Azure Resource with a project code that’d be the identifier of the associated project you’re running. Having a granular way of calculating costs allows you to come back to your stakeholders with what they are consuming and potentially charge them back.
Limitations of existing solutions
Currently, in Cloudyn, not all tags are brought back in the UI as it seems that only tags associated to VMs are surfaced, which is far from representing all kind of costs incurred by your activities, although VMs are indeed costly resources. With the Power BI app, tags are there but on their original form, meaning an arbitrary array of tags for each tagged resource. I say arbitrary as some tags are added by Azure itself. Therefore, it is very hard if not impossible to exploit this in reports, even when using Advanced Filtering. Continue reading
For this episode, I have created another chatbot that is aimed at helping factory workers to intervene on machines whenever they encounter operating problems. This factory comes with a specific jargon and workers are surrounded by permanent noise which can obfuscate worker statements when they give vocal orders. We’ll tackle these constraints by leveraging the Custom Speech service with the bot framework. We’ll also see how Custom Speech differs from Speech Priming that I talked about in episode 8.
If you’re not yet familiar with the bot framework and the cognitive services in general, I strongly advise you to watch my other episodes as I will only focus on Custom Speech and I will not explain things I have already explained in the previous episodes.
Now that we built a chatbot using most of the NLP-related APIs and that we saw how to categorize incidents based on end users screenshots thanks to the custom vision service, it’s time to see how to add speech to this bot! We’ll see several flavors of speech services and we’ll see how to fine tune speech-enabled bots with speech priming.
You can watch this episode on Channel9 https://channel9.msdn.com/Blogs/MVP-Azure/Cognitive-Services-Episode-8-Leveraging-speech-services-with-chatbots
Microsoft recently announced Azure Managed Service Identity (MSI) which in a nutshell, is a way to avoid storing credentials in code or in locations such as the web.config, the app service settings etc…thanks to an automatically provisoned Service Principal (bootstrap identity) that you can leverage using the App Service (or other components supporting MSI).
As Microsoft highlights in the above article, even Azure Key Vault didn’t really solve the problem of disclosing credentials since your code needed credentials to get access to the Vault. Therefore, any developer could have written a console app, connect & retrieve the actual secret values from the Vault.
Recently, I wrote a short blog post on how to provision Azure Active Directory (AAD) Apps in a highly controlled way, so I will not repeat all I said there, but it a nutshell, the idea is to make sure DevOps can automate the creation/update/deletion of AAD Apps entirely from VSTS while not being able to interact with non-DevOps apps.
Here is a step by step process on how to get there. Note that almost everything could be done from VSTS but, often, in organizations, the below tasks will involve different people & even different teams, hence the reason I decouple all the tasks. Continue reading
Besides promoting a new collaboration mindset between development & operations, DevOps’ primary goal is to use tooling in order to reach continuous development as well as continuous deployment. As it implies a cultural change, it often cristalizes tensions between the involved stakeholders but I’m not gonna debate about its current effectivness and reality within the enterprise, instead, I’m going to focus on automated deployments of Azure Active Directory Applications.
With the recent publishing on my 6th episode, I just closed the chapter on using NLP with Azure Cognitive Services. In this course, I explain little by little how to build a chatbot that deals with various tasks, each task being associated to one of the Cognitive Services.
As the NLP chapter is closing, here is a recap of what I covered so far:
Episode 1 In this episode, I will draw the AI landscape of the Microsoft ecosystem. I want you to be a little more familiar with fundamental topics such as Machine Learning, Deep Learning and Natural Language Processing which might sound a little bit confusing for many developers. Once the high-level concepts will be covered, I’ll make an introduction of the Azure Cognitive Services and I’ll try to quickly answer the “what’s in it for me” question out of real world examples mapped to the various services. If you’re a hardcore developer, you might be disappointed by this episode as I will not show code yet, but by the end of it, you should understand when to use what and how to manage customer expectations. For the “how to bits”, I invite you to join me at Episode 2. Continue reading