MS, Tableau and AWS working together IAM; BigQuery bias; Cosmos DB and vector search

AWS
The AWS blog post explains how to integrate Tableau and Microsoft Entra ID (formerly Azure AD) with Amazon Redshift using AWS IAM Identity Center for single sign-on (SSO).

The process involves configuring IAM Identity Center to manage user access, setting up Tableau to use SAML for authentication, and integrating with Amazon Redshift for secure access to data.

An incredibly long post written by 7 people. Should be a good, technical read.

GCP
Of course I expect it to be biased in favour of BigQuery.

The whitepaper highlights the economic benefits of using Google Cloud's BigQuery compared to other data and AI platforms.

It emphasizes BigQuery’s ability to scale automatically, reduce operational costs, and speed up time-to-value.

That’s just the summary. I only imagine the praise for BigQuery. I’ll read it nonetheless.

AZURE

An introduction to Azure Cosmos DB's vector search capability using DiskANN, a scalable approximate nearest neighbor algorithm.

This capability enables efficient and accurate searches over large datasets, enhancing applications like recommendation systems and semantic search.

The post also discusses how Cosmos DB integrates this feature to handle vast amounts of data.

I love this new stuff. That’s why I write about it.

DATABRICKS
Discussed is an end-to-end approach to data intelligence using Azure Databricks and Microsoft Fabric.

The post covers data ingestion, storage, transformation, machine learning, and business intelligence processes.

The architecture integrates multiple Azure services such as Data Lake, Synapse, and Power BI

Take my word for it. There is a very nice looking architecture diagram in the post.

SNOWFLAKE
Trust me when I tell you I’m trying to find tech news on Snowflake. Even the almighty Seeking Alpha is waxing poetic on the stock…

GRAPH DATABASES
Mr. motorcycle jacket (I have names for all major CEOs), delves into the world of Graph Databases.

NVIDIA and ArangoDB have collaborated to enhance NetworkX performance for graph analytics.

By integrating NVIDIA's cuGraph, users can accelerate NetworkX with GPUs without code changes, boosting performance by up to 600x. ArangoDB offers a persistence layer for scaling and data storage, enabling seamless data manipulation and analysis.

VECTOR DATABASES
You know AWS has mountain ranges of money when they can produce scientific paper level blog posts at scale.

An excellent read on how to enhance knowledge retrieval in Retrieval-Augmented Generation (RAG) using Amazon OpenSearch by integrating both sparse and dense vectors.

Sparse vectors are used for keyword-based search, while dense vectors improve semantic search by capturing contextual relationships.

This combination improves search accuracy and relevance, enabling more effective knowledge retrieval for machine learning models, such as those used in natural language processing tasks.

DEEP DIVE
This deep dive is about two different subjects. You’ll get the gist of it as you read on. I think that is what you call foreshadowing, I think.

I met up with a former co-worker of mine that I worked with for almost 13 years, after work yesterday. He is a very good-natured guy, smart, and insightful. I would only add that he should not be crossed. Management felt his wrath because they would come up with a lot of asinine ideas. Management would get an earful from me too. They labeled us “the trouble makers”, and that is a badge I wear quite proudly.

We had to fight with management side-by-side and on our own on many occasions. The reason why we were so adamant about our stances is that we would back our claims with cogent, irrefutable evidence that could not be disputed.

We did not fall for buzzwords, latest trends, and any other foolishness.

What I am getting at is twofold. The first point is that you have to be good at what you do and know your technology inside out. Staying up to date is key, hence this newsletter. The second is maintaining your network. You never know when you may need a helping hand, or a new job for that matter.

Just a few good people in your circle can go a long way. We were always compering notes, and coordinating with each other so that dumb decisions would not be made by management.

Once you back that up with your experience, curiosity, skills, and logical thinking as the bedrock, along with a good network, you can last in the in this realm called IT.

Gladstone