Introduction
This story might sound like I’m making excuses, but I am really just giving an explanation of the story behind a recent interview I conducted. I want to take this opportunity to clarify or expand on some of the topics we discussed.
The Podcast Interview
I recorded an interview with Brian Oblinger and Garth Miles of Alteryx on Friday, March 9th, 2018. That interview was published today as Episode 2 of the Alter Everything Podcast.
In this interview, we discussed a variety of topics including backgrounds, work histories, big data and the role of Alteryx in data science projects. As I listened to the completed interview, a few things occurred to me that I’d like to clarify and/or expand upon in this article.
I Was Recovering From the Flu
When I wrote this article on February 26th, little did I realize that I’d be knocked out for a couple of weeks with the flu. While my wife was cruising in the Caribbean, I came into contact with the flu and began taking a bruising. By the time I saw the doctor on Feb 28th, my fever was approaching 104 and my brain was starting to sizzle. All I could hear was static, and I was barely ticking along. This was the night I was diagnosed with the Type-A flu and barely escaped going to the ER.
After a couple of shots in each gluteus maximus muscle and being given a variety of other medications, I began the journey to recovery. To say that was a pleasant time in my life would be a misnomer. I had a tough week of isolation, fevers, chills, and a “pain band” that went around my midsection, just around the lower area of my ribs. This pain band made me want to stop breathing because it felt like my body was wrapped with giant zip ties and every breath was agonizing when the pain band was active.
By March 9th (9 days later), I was feeling well enough to do the interview, but I was sure that I was still mentally a little slow. By taking some cough medicine, I was able to give the interview without too many coughing interruptions or major screw-ups.
Topics That I Want To Clarify and/or Expand
When I listened to the interview today, I realized that did not give concise and complete answers to a few of the questions given to me. I feel disappointed that I let Brian and Garth down at certain points of our time together. This likely occurred because my brain was not functioning at peak capacity at the time due to the lingering illness.
This article includes some things I would like to add to the discussion to give more complete answers to their questions. Without a lot of practice giving interviews, it is really hard to say everything that is important on a particular topic when the microphone is recording every spoken word.
Topic 1 – The Combination of Alteryx and Tableau
To begin, I should have said at the very beginning of the podcast that the combination of Alteryx and Tableau is a very powerful platform for solving data science problems. I know this to be true because of my programming experience and work history. I spent over three decades writing custom computer codes to solve challenging data science problems. In many cases, custom data sets had to be created by blending data from multiple sources.
The production of those custom data sets happens because of tools like Alteryx and Tableau, as well as a lot of practice in doing the work. It also happens because I can pull knowledge and experiences from math, natural sciences, statistics, programming, logic, quantitative experiences, photography, and a multitude of other subjects and combine this information to solve problems. For this reason, I have a passion and deep appreciation for these software packages. These two companies and their products have allowed me to reach my potential as a quantitative worker and to explore data without limits.
This passion has motivated me over the past five years to write over 320 articles explaining why the combination of Alteryx and Tableau is so powerful. I also knew that there is no way I can fully explain my insights in a 45-minute interview. This is why I am writing this article.
Additionally, I practice this craft of solving problems in advanced analytics every day on my job and some of the things I have accomplished are truly inspiring and unbelievable in some ways, even if I can’t share these results. The speed and comprehensive nature of what I can do to data using this tool combination are historically unprecedented.
I know the truth of what this software combination can do, and I should have stated my most important conclusion first. I feel like I missed a major opportunity to promote Alteryx and Tableau as a formidable data science problem-solving platform.
Topic 2 – Preparing For A Data Science Career
When Brian asked me about preparing to pursue a career in data science, I said that getting hands-on with data was very important. Working with a wide variety of data is essential. Those are both true statements.
I should have added the importance of pure academic studies in mathematics and computer science. Having a strong quantitative background is very helpful in this career. So the next time you hear your kid say that “I don’t care about this math problem because I’ll never use it in my lifetime”, be sure to tell them that they will use it more than they can imagine if they get into the field of analytics!
In my case, I took math courses until the point of mental failure, even though I was only required to take one calculus course for my undergraduate degree. Courses in linear algebra, vector analysis, numerical analysis and numerical calculus are all great to take to prepare for data science. If computational work is of interest, math courses that specialize in developing and understanding computational algorithms are good choices.
In other words, encourage our students to never stop taking math while they are in school. It forms the foundation of all topics in data science. Mathematical concepts are intimately linked to one another and allow you to perform a wider variety of studies in advanced analytics.
Topic 3 – Diversity in Computer Programming
Learning how to program in multiple computer languages is also very helpful in the career of data science. Learning object-oriented design is an important topic to include, too. Since computing languages are constantly evolving, be sure to pick the ones that are emerging, rather than the ones that are falling out of favor. In 2018, Python would be a good choice to learn. If you are statistically-oriented, taking courses in R programming is a good idea.
When Garth asked me about what languages I have used, I neglected to say basic, visual basic, and VBA. One surprisingly powerful platform I have used is VBA within Excel. I have written technically sophisticated programs larger than 30,000 lines in VBA. VBA is a great starter language for data science.
Finally, it goes without saying that learning Alteryx is absolutely the key to a successful career. Since Alteryx covers such a wide spectrum of topics from data prep, to geospatial analysis, to demographic analysis, to descriptive, predictive and prescriptive analytics, this is the tool of choice!
Topic 4 – Defining a Data Scientist
When Brian asked me to define a data scientist, in the original interview I believe that I gave him an explanation that didn’t make it into the final podcast due to time considerations. One of the things I said about being a data scientist relates to having a wide and diverse technical skill set.
I believe that every “data scientist” should be able to complete the full spectrum of tasks that occur in an analytics study. This means being able to (at a minimum):
- access data via local and networked systems
- perform quality assurance/quality control (QA/QC) checks on the data
- properly prepare, structure and focus the data for analysis
- complete the analysis
- create visual and quantitative results
- explain the results so that non-technical people can understand the insights that the study uncovered.
Each of these tasks requires specialized skills and each of them are important. It takes years of practice to develop these skills in their entirety. Also, learning how to design data collection activities is a great skill to have because it forces you to understand what information will be needed to answer the business or science question being posed. To read about my complete perspective on improving your ability to work with data and to improve data comprehension, please read this series of articles.
I also thought that Garth’s answer to that question was excellent. I especially liked his comment that not everyone needs to be a “data scientist”. Being good with data preparation and custom data set creation is an excellent and much-needed skill set. Learning to work with a tool like Alteryx combined with a great visualization platform like Tableau will give you superbly marketable skills. Not everyone needs to run machine learning algorithms. Learning to work with deterministic data by itself is a great skill to have will allow you to have an interesting and long-lasting career.
Topic 5 – Future Explanations
Finally, by the time the interview ended, I felt like I really hadn’t said that much. We originally were going to talk about what I learned from the climate studies I performed. We only glossed over that.
What I realized is that it is hard to compress over 30 years of continuous work experience down into a 45-minute interview. There were important statements I made about Alteryx that I didn’t have time to fully explain. For a few of those statements, I’m going to define and document them in upcoming articles. If you are interested in reading more about those topics, please consider subscribing to this blog.
Final Thoughts
Here are some additional articles related to the topics I discussed in the interview. Thanks for reading!
Pingback: How To Achieve Rapid Analytical Insights Via #Alteryx and #Tableau | Data Blends