Time to bid SAS, Goodnight!

Victor Hugo once said - 'No force on earth can stop an idea whose time has come'. To all fellow professionals who are either practitioners of Machine Learning (Analytics/Data Science/AI) or consumers of products or services that are propelled by Machine Learning, would care to note this profound philosophy, propounded by one of the greatest French writers of all time.

To turn the headline of this article on its head, which by the way, was my original intention, would be to re-read it as follows – “KNIME to bid SAS, Goodnight (pun intended, for all those who got it).

Subsequent to my earlier article published this week (https://www.dhirubhai.net/pulse/analytics-industrial-revolution-from-occult-ordinary-mukherjee/), a large number of readers, asked me for a definitive verdict on the tools which I had listed as being at the forefront of the Analytics/Machine Learning Industrial Revolution. They wished to know, which amongst Alteryx, RapidMiner and KNIME was the first amongst equals.

I have used both RapidMiner and KNIME and have watched Alteryx in action. The platform or workbench that takes the cake, has to be KNIME, and in my honest opinion, is the king among these tools. KNIME has the added advantage of being an open source platform (though it has server and cloud versions which are supported and can be bought/subscribed to, as per need).

My introduction to KNIME was as sudden and serendipitous as it can get. I came to know through a colleague of mine that one particular project team was in a last-minute crisis (as it often happens in time crunched projects) and was not being able to get a reasonably good predictive model and this particular gentleman appeared like a messiah and built ~300 models in about half an hour and collated all the results from the model runs, and was also able to give 5 very good models to the team to choose from. And the word went around in the grapevine that there was a secret weapon which we had used. And that secret weapon was called KNIME!

This was two years back and every time after that, whenever I have opened up KNIME to solve problems of data fetching, data processing, model building and model validation (machine learning) and charting out results I have stopped my train of thoughts to thank this person in my mind. KNIME has been such a revelation to me and I have used the tool with the unbridled joy of a child who has been provided with a wonder toy.

KNIME has something to offer to everyone.

·      To the accomplished coder it provides incredulity (and a sense of dejection that this workbench was not available in the earlier days) that what can take days and weeks to code, can now be achieved in a matter of minutes and hours. (I used to be an expert in my earlier career in C, SAS, SQL, Pascal etc.)

·      To folks like me and others in the senior management of companies, it provides a fresh lease of life to be able to get to the thick of things once again and claim back your right to a hard skill that you once had, but lost along the way, while you were busy climbing corporate ladders, selling to clients, managing teams and people and strategizing on the next big things that your firm should do. These are all great things we are doing, but it indeed is a great feeling to be back in the trenches with your soldiers with this new bazooka – go give it a try! And for a change, you can lead by example. If you are telling your teams to become more efficient, using buzzwords like better, faster, cheaper, here is a chance to show them the way-start “KNIMING” yourself. As the great Mohandas Karamchand Gandhi said, “Be the change that you wish to see in the world”.

·      To the business user it provides a wonderful opportunity to apply Machine Learning to his or her own business and reap immediate benefits, without having to learn how to code

·      To the data scientists, AI pundits and Analysts – it gives you the chance to focus most of your time on gleaning insights from the data/models to help your clients/teams/businesses with better decision making by freeing up the humongous time that is spent in writing reams and reams of code to do endless data cleaning and data processing/manipulations and model iterations. That is the worst kept secret of the Data scientists – most of the time is spent on data cleaning, data manipulation before it is even ready for building machine learning models or performing statistical analysis.

My weekend job is that of an Adjunct Professor of Marketing Analytics at the Great Lakes Institute of Management and every time, I introduce my students to this tool, I sense the catch22 feeling amongst them. And many are brutally honest and vocal about it – “The industry is still hiring folks with SAS, R or Python skills, so how would KNIME help me “. To them, I only have one thing to say, something which the greatest ice hockey player of all time, Wayne Gretzky revealed as the secret of his success on the playing field – “I skate to where the puck is going to be, not where it has been.” By the way, they don’t teach this to you at Harvard Business School (with due respect to Mark H McCormack and What they don’t teach you at Harvard Business School).

The naysayers and doubting Thomases, you will find aplenty. Some of them are so scared that they would not even touch the tool, lest their worst fears of what is being said about the tool by people who have used it, comes true. They would very much stick to their good old tools and close their eyes like a frog in a pond and believe that somehow the plague of industrialization and automation, will leave them untouched. To quote the Yiddish proverb made popular by Malcom Gladwell – “To a worm in horseradish, the world is horseradish”.

As I was driving to office in the morning today, my favorite radio channel was belting out this number by Roscoe Dash – “And we gon' have a good (good) night , Good (good) night “ almost as if by telepathy, it knew that I was going to pen this article today. I did recall the great many projects that I have delivered for clients over the years using SAS, but the honest portion of my mind must not deny the brutal fact that the “Time for KNIME has surely come”.

#datascience #machinelearning #artificialintelligence #ai #ml

PS: Sorry to leave in “such a hash” – but am told that this helps reach a larger audience!

PPS: The opinions expressed by the author are his very own and doesn’t represent any organization or company that he is/ has been a part of.