Mike Allen Systems

Doing Smarter Things With IT

Offering more cloud solutions

I’ve been ramping up my BI offering, centered around PowerBI. In order to offer my clients a more complete solution I have become a cloud solution provider (CSP), partnered with rhipe.com.au. This enables me to offer a range of Microsoft products including Office 365, Power Bi Pro and Azure services. I’m still adjusting my offering, but, if it’s Microsoft, I can supply it to you at a reasonable price and offer some excellent support.

Focusing on Business Intelligence and Power BI

Much of my work over the last few years has focused on Data,Reporting and ETL, all components of Business Intelligence, BI. Now that Power BI is establishing itself as a BI market leader I find that I can re-focus my business more easily on BI and I intend to promote and market Power BI, and my abilities and experience in Power BI. I have been pursuing this in two areas, I am creating BI how-to videos on YouTube , and I have a new Mike Allen Systems web site to again focus purely on BI. Check out my videos and new web site, but most importantly check out Power BI , it’s a free download!

Microsoft Professional Program for Data Science

I recently completed the new Microsoft Professional Program Certificate in Data Science. I've a background in software and data migration, and I knew how to use the tools. The course was intended to formalise my skills. Well, was I in for a surprise.

I'd already completed the Power BI Course, and first up was Querying Data with SQL, what could possibly go wrong? The courses were through edX and were mainly USD $50, they have all gone up to $100 since then.

The first speed bump was choosing a statistical language, I flipped a coin between R and Python a few times, and the gods of statistics seemed to favour R. There was also a new plug-in for R in Visual Studio, which made life easier. I kind of fell in love with R, anyone who likes to manipulate a whole dataset at a time would be at home with this language. My favourite bit of code just extracts the year from a date column and adds it as a new column on the end, cool! The density graphs are an absolute delight, as are all the graphical functions. There are R packages for everything and a vibrant user community, this is all good. I also found R useful for generating test data.

Next up was learning some real Statistical Thinking with ColumbiaX. I had done my stats training years ago while studying Applied Science. This course cleaned out all the cobwebs and got me enthused, there are more stats courses from ColumbiaX which I may get into in the future. Great stuff to get one Statistically Thinking. They've taken this off the Data Science track now, I can see why, but it needs a suitable replacement.

Then things got a little more technical with the Principles of Machine Learning. I'll never look at a regression graph in the same way. the introduction to Azure Machine Learning was a real epiphany of drag and drop Data Science. This is amazing stuff and right at the leading edge of cloud technology. I've has some exposure to Azure but this was really something else. Although calling an Azure Web Service from Excel did give me a touch of cognitive dissonance. Another great feature I found in Azure was Jupyter Notebook, which is a lovely open source intelligent notebook that can execute code. Wow! Executable notes!

I think every design team could benefit from this, what a way to prototype data transformations.

My decision to do the Developing Intelligent Application was a little off track, but I bought a book on C# to help me. This is part of the Applied Data Science section and there are some more data oriented courses available. Developing a bot was fun, and there was lot in this course. Analysing Twitter Feeds was great and bought in the Azure IoT hub and other Azure elements. This is obviously very popular and you can do all this with Microsoft Flow now :-(. Lot's more technical stuff in this one, I've even made friends with JSON. I did discover that SQL is useful for data analytics, the old is made new again!

The final Capstone project was a return to data analysis and Machine Learning. The hardest bit was writing the Data Analysis report and reviewing other students’ efforts.

This Certification took me over six months to complete and I learned a lot along the way. I paid for all the courses to get certification, however casual auditing of the course is available. The Professional Program offers a good conceptual learning alternative for those not seeking Microsoft technical certification. I suppose the down side is that you are expected to do a lot of technical things, like C#, and get no real certification for it.

I can't speak more highly of edX. Their courses are so well organised, the platform is easy to use and the content is excellent. Microsoft have made a great choice in partnering with them.

Was it all worth it? Absolutely! I'll be back to edX as soon as I've finished my book on Bayesian Analysis :-).

Bell and Whistle V2.0 – bigger and better

I was reasonably happy with my first Bell and Whistle data set, but I felt it needed more data, a lot more. I’ve just done a course on using R for simulations, so I thought I would generate some test transactions with ‘days’ stretching over 10 years and a variety of customer and product numbers. I settled on 113 Customers and 50 Products and used a randomized section of a normal distribution to give me around 50,0000 transactions that became denser over time. That’s a complicated way of saying sales increased over time.

What’s it for? Primarily for demonstrating data visualizations in products like Excel, PowerBI and whatever takes you fancy. I’ve been using the previous version to produce some training videos https://www.youtube.com/channel/UCtvirMVp8r8j6jGbWE8SqIA

I’ve decided to distribute Bell and Whistle V2.0 as an Access Database and as a Excel workbook that uses the data in a Data Model. You’ll need Excel Pro to fully realize the work book. The Access database is just data, with the table relationships. The Excel data model has some DAX and Measures.

These file are available on Google Drive try, https://drive.google.com/open?id=0B9nJpvPnppwldm9tWU9GbV9CT2s for the Access Database and https://drive.google.com/open?id=0B9nJpvPnppwlcy1WRW9hRDZRUG8 for the Excel Workbook.

Have a play, let me know if you like it.

Project 21–A more efficient IT methodology

For quite some time now I’ve been working on an IT development methodology called ‘Project 21’. This has emerged from a number of medium sized projects where no methodology was in place and a lightweight pragmatic approach was required. I’ve just migrated the web site to a new host and the latest version of WordPress. This year I’m going to do some more work on it, particularly to address the development phase and the implications of Agile on this sort of methodology. Project 21 is designed to be used and does not attempt to comply with any particular standard. Unfortunately standards compliance would be self-defeating as this methodology is designed to be minimalistic and something that can be tailored to individual project requirements. This is the Project 21 web site. Please feel free to use this content, any attribution or feedback would be more than welcome.

Bell & Whistle Dataset

In the past I’ve often demonstrated concepts to clients using their own data, recently I have been working on a dataset using my own data so the I can share demonstration and training videos and other materials more readily.

My first effort was a training video for pivot tables, although it ended up being more about how to import CSV data into Excel!

 

The current state of the data is here in a Access Database BellAndWhistleV1.2

This data is supplied free and with no guarantee, download it, use it, attribute if appropriate.

Analysing Data Ownership

Finding data ownership can be a really worthwhile task. There you are with the task of documenting a database, someone suggests that you google a few definitions and and do a bit of cut and paste. Then you decide to do something worthwhile, and determine who owns the data, this may cause a few sharp intakes of breath from those that think the IT department owns the data. If you want to attach a bit of methodology, think of applying a Zachman Framework to the data structure as a business model (don’t get too hung up on Zachman). Once you’ve started identifying ownership you can start thinking about dependencies, permissions, life-cycle’s and all sorts of other data goodness. That list of data owners will be useful for a number of roles, managers, analysts, testers etc. The structure may not be simple, you may find records with multiple owners and have to negotiate who is the master, and it’s a good idea to have a single master, for transactional data it may be you!

This is where the a governance model comes into play, IT can be the keeper of the metadata, assuming you have the luxury of an IT department else the responsibility falls to the data owner. The owner will have contributors and consumers and needs to be aware of who they are and what their rights are. Individual data items may be owned by customers, please consider confidentiality and privacy considerations. Useful information can be collected down to the individual data item, you may need to define some domain information for some data items, alternate names and scope.

Once all this information is is collected, or maybe even sooner, it is important to publish, and publish so that all stakeholders can access the information. Then when someone says “We need a new keyword for this!”, they can find the ownership and responsibilities of “keyword” and act accordingly. Database documentation, if you build it well, it will be adopted and maintained by the business as an asset.

A New Blog, a Fresh Start

In the beginning I was in love with WordPress and how simple it was. When I wanted to host my own blog, it got more complicated but it was bearable. I started using ASPX and C# a little, but I stuck with WordPress as there was no equivalent C# available. We used ScrewTurn an ASPX, C# product at work, I selected it, but then it was discontinued. Then I had a few issues with WordPress and I thought I’d look around and BlogEngine caught my eye, ASPX, C# Open Source, fresh and fast, what wasn’t there to love. Even better Open Live Writer became available on the Windows store, a match made in code,

So here we are with a fresh blog and a new blogging client, I’ve cleared out some old content and I’m ready to blog again.

Meanderings on Privacy

Privacy is often treated as a legal obligation for a business . Privacy is more than a legal obligation, it is at the heart of the relationship between the business and its customers. Confidentiality and the customer expectations of confidentiality must be aligned in order to build trust in customer relationships. Privacy policy is often regarded as a requirement to comply with government regulation rather than an aid to better customer relations.
Privacy requires thought about a business and its interaction with external entities. Let me pose a simple questions. Should the privacy policy of a website and the organisation that owns the website be the same policy? How does all of this relate to terms and conditions of use of a web site? Sorry that was another question that came to mind. What about web site cookies?
How should privacy policy be presented to external entities, individuals, Companies and Agencies? A summary statement would seem to be a good idea (please don’t include the words ‘we take your privacy seriously’, devalued from over-use), a good summary should incorporate your business approach to privacy, particularly important if you have many individual customers who have rights under the privacy act. Various parts of the policy can then address individuals, other businesses and agencies and areas that cross all of these such as data retention.
Another area that is often ignored is paper records, most businesses of any size acquire paper documents. If you are lucky you have retention and disposal policies for paper records, if not they should reflect your electronic standards. But that’s a whole other area; however it does need to be considered in your privacy statement.
Of course this is just my opinion, but next time you look at your privacy policy, just think, what is this doing for our business? have we looked at this recently?