Skip to main content


Azure is microsofts cloud service. It is like AWS, but only more microsofty. I have worked extensievley with Azure and find it to be a bit cumbersome, with rough edges. Understanding the CLI was key for my development, made things so much easier. I think this experience is something that is shared across cloud providers.


Cognitive search is like Google search on the data you have available in a storage container or database. Reasonably easy tutorial on cognitive search

Cognitive search is based on indexing the documents in the database, a good overview of the search index is available from the Azure documentation

Using the command line interface (CLI) is always a good way of understanding and learning the services. The documentation for cognitive search CLI is pretty good. In my experience doing continuous integration (CI) and continuous delivery (CD) makes knowing the CLI mandatory.

Make better indexers with Powerskills Repo

![[../../static/images/Pasted image 20230615144221.png]]

Azure Cognitive Search's Semantic Search is a feature that enhances traditional search methods by understanding the intent and contextual meaning of search queries. By leveraging AI and natural language processing, it can deliver more relevant search results, even when queries are phrased in a conversational manner or have ambiguous terms. It can also provide advanced query capabilities, like document summarisation, named entity recognition, and personalized ranking of search results.

Demo using Cognitive Search and ChatGPT

Machine Learning Studio

Document intelligence.

Document intelligence is usually scanning documents such as pdf for their content using OCR (Optical Character Recognition). It can retrieve a lot of information and has a very handy feature to extract tables and key value pair. I used it to create quite powerful features with LLMs. For example handling enterprise emails and requests it is quite powerful.


ML-Ops in Azure Examples

Infrastructure as Code




  • SKU is the product tier of the resource in Azure. In Azure, the term "SKU," which stands for "Stock Keeping Unit," refers to a specific version or configuration of an Azure service or product.
  • Purge protection in the key vault is that you don't delete everything by accident. Postgres backups are, for example, linked to the Postgres resource.