Are you already profiting from the No-Code megatrend ?
Where some see python, most people see hieroglyphs. Not surprisingly, data science seems as far away as Egypt, but this is changing. With the emergence of the cloud and technology (e.g. memory, storage, ..), IT resources are cheap and easily available. A laptop & internet is all you need.
What was holding us back?
The scarcity of development resources. If I don’t know web development, I couldn’t build a (nice) homepage. This is why “No Code” applications are a gamechanger: Those allow us to solve computer problems with drag & drop menus. While some applications are already mature (e.g. building an online shop with shopify), there are still primarily known by industry insiders (e.g. google autoML, github code templates).
Why should you care?
Business Leaders now have an abundance of (new) tools available. Technology is moving from process (IT infrastructure) to product (“smart” solutions). No Code overcomes technical limitations and allows you to focus on the problem, which accelerates productivity.
But do I get what I need? And why would they need me?
Truth be told, it is unlikely that you find a plug & play solution. Every model template has to customized to your dataset, a Python intro course (e.g. coursera) and intermediate googling skills are helpful. Then again, common problems (e.g. price prediction, any form of classification, natural language processing) are fairly easy to tailor to your cause (e.g. code-along resources on medium).
What is much more important is A) a good understanding on your data (what is it your predicting? Which variables are factors? Do you have all relevant variables in your data) and B) limitations of the chosen model. Getting those two right is crucial and will make or break your results.
No-Code data science is like buying pasta instead of knead dough: It avoids hassle & a great way of getting started, but if you don’t have a sense for your ingredients (data) and sauce (business context), you are better of with take out all along.
This is why expertise, experience and abstract thinking are (still) key to success and irreplaceable for any data science project. There is a reason why people talk about “AI fatigue” (more on this at a later date), as many projects didn’t meet expectations.
In my opinion, this is why I am a fan of no-code: It enables non-techies that bring essential skills to the table. Adding product experts & business leaders to the conversation, adds diversity allowing for more holistic solutions.