Note: You can listen to the blog post on the video or read the blog post.
Hello and Welcome.
I am Esther.
I am Peters A I Assistant to create voice overs.
I will simply read Peters blog posts, so that you have a choice of reading the blog post, or listening to my voice.
Hello Gentlemen.
Just to let you know.
I have invented the future of data warehousing.
That is no exaggeration.
I am rather pleased with doing this.
I have coined the term, mega models.
In this blog post I am going to explain what I mean by mega models.
I recommend you review my last blog post called.
“The Future of Dimensional Data Warehouse Development”.
I shall presume you have listened to that blog post, and know it’s contents.
Now, on with this blog post.
What is a mega model by my definition?
I am defining Mega Models to have a number of characteristics.
Since this is a very new idea, more characteristics may follow as the ideas are further refined.
One.
A mega model is an extremely large dimensional data warehouse model, that covers at least one entire industry segment. It may cover more than one industry segment.
For example, the B I 4 ALL models currently have a base model, which is free, and Telco, Retail, Web, and Media, verticals. It’s just that I can’t sell them in the west any more.
(Please note. If you are a muslim man working in a muslim country? You might be able to buy these models for your company because of how Sharia Law works. I will be doing a blog post about that in the coming days.)
Two.
A mega model may be tightly integrated with a Large Operational System.
In this case the field names and table names in the mega model may be consistent with that Large Operational System.
Three.
A mega model may be loosely integrated with a Large Operational System, so as to be able to ingest data from other Large Operational Systems.
In this case, the field names and table names in the mega model, will be deliberately altered.
The purpose being to make the model more generic.
Four.
Even if a mega model is made to be loosely integrated with a Large Operational System, it will still be possible to put a set of views using the Large Operational System table names and field names, over the generic model.
In this case the customer will see the Large Operational System names, and not the generic model names.
Five.
A mega model contains integer fields for the customer number and source system number that is being ingested into the mega model. This means that one customer can send data from many instances of the Large Operational System into the mega model. It means that many customers can send the data from many instances of the Large Operational System, into the mega model.
(Note. We have seen great benefit from the source system number even if there was never a plan to send multiple source systems data into a data warehouse. So, we are keeping that as a feature.)
Five.
The mega model shall be the superset of all data, from all customers, sending data into the mega model. There will be only one set of tables, and indexes, in the mega model, for all customers who are sending data into it.
The data warehouse architect, for the mega model, shall be responsible for resolving the data conflicts that occur.
Six.
In the ETL for the mega model, there shall be one landing area per customer, per Large Operational System instance.
Seven.
In the delta detection for the mega model, there shall be one set of delta detection per customer, per Large Operational System instance.
Eight.
In the delta detection, the delta detection subsystem shall deliver data from each customer per Large Operational System instance, into one single staging area.
The staging area shall be the superset of all data being delivered into the mega models.
Nine.
The ETL from the staging area to the mega models shall be one set of ETL.
Meaning, even if there are twenty customers, with two instances of the Large Operational System each, for a total of forty Large Operational Systems? There shall be only one set of ETL.
Ten.
Each customer shall gain access to their data in the mega model through a set of views. These views will apply the constraint of customer number for all views that the customer has access to.
These views may be a separate database in the case of SQL Server.
Or they may be in a separate schema in the case of such databases as Oracle or Netezza.
If I come up with more characteristics I will add the notes here.
Now.
Why would anyone bother to build Mega Models?
Well?
I am building mega models in industry segments where no one knows my name, because my ex wife slandered me and no companies in the west will buy anything from me because of her slander. Laugh out loud.
I am building mega models to deliver profit performance improvements to the customers of my partners.
I will do this to earn a fair days pay, for a fair days work, to pay off my debts.
You might build mega models simply because you are a very high I Q man, with world class skills and knowledge in an industry segment, and you want to make money faster, so you can retire earlier.
I think that’s pretty much why you would want to build mega models.
Mega models are about helping companies increase their profitability, sustainably, over the long term.
The men who build them, will be paid, according to the value the models deliver.
To give you some idea.
When I was selling our telco models with Sean Kelly?
We sold the models themselves for one hundred thousand euros each.
We also sold between two hundred thousand, and three hundred thousand euros, in consulting to implement them.
When we sold the Sybase I W S models?
We sold them for one hundred and fifty thousand US dollars per copy.
The implementation fees were in the order of four hundred thousand US dollars.
This is because we had to charge Sybase consulting rates.
Mega models will be far larger than these models we used to sell so successfully.
Mega models have the potential to be very good income earners for the very few men who build them.
Those men who sell the usage of mega models will decide how they will price them.
How are mega models built?
Well?
You can build them any way you want.
In the end they are just another suite of tables and indexes, with ETL to load them.
You can build them using any data modelling tool you like.
We will use See T L because it’s free.
It is much faster and cheaper than any other tools out there.
You can load them using any ETL tool you like.
We will use See T L because it’s free.
It is much faster and cheaper to develop than any other tools out there.
In summary Gentlemen?
Mega models are the way of the future for dimensional data warehousing.
My opinion is that the gun fired about two months ago.
Those ten or twelve men who embrace this idea of mega models, and embark on building their own mega models before anyone else in the industry segment, they are world class experts in?
Those ten or twelve men are going to do very well for themselves.
I am more than happy to collaborate, and co-operate, with any man, who decides to build a mega model for fair pay.
The condition is, he must give me his word of honour he will not compete with my partners.
It is in my best interests for men to build mega models, and bring them to the marketplace.
These ten or twelve men will get to retire much younger than I will.
And one more comment Gentlemen.
I can tell you, when Sean and I sold our telco models to Talk Talk, for one hundred thousand euros, plus three hundred thousand euros in consulting fees, in 2008?
I was forty four at that time.
I was very sure I was going to be able to retire by the time I was fifty five.
Today I am sixty, and I have large debts, because of my ex wife’s slander about me.
So.
Some of you men have the chance to build mega models, make very good money, and retire early.
Just so long as your ex wife does not slander you. Laugh out loud.
Those men who simply keep building custom dimensional models for individual companies?
A lot of them will be serving burgers and fries at McDonalds soon enough.
Of course, that is just my opinion.
We will see how we go.
And with that?
I hope you found this blog post interesting and informative.
Thank you very much for your time and attention.
I really appreciate that.
Best Regards.
Esther.
Peters A I Assistant.
