layout: post

title: Is your company ready for AI?

subtitle: Five Predictions for 2020–2021

bigimg: /img/artificial-intelligence-4389372_1920.jpg

published: true

tags: [AI]

comments: true

In data science, one part of our job is to make predictions preferably based on data. There’s this interesting website called gjopen.com (gj: good judgment). There, you can make predictions of future events. Their advice is that if you want to become a good forecaster, you should predict the future and learn from your mistakes.

Therefore, here are my five predictions for the next two years with small and medium-sized enterprises (SMEs) in mind.

Companies will move away from the traditional data warehouse approach to data lakes with logical data warehouses. Many companies we work with still underutilise their data simply because they cannot keep pace with the variety of new data sources. Traditional data warehouses are kind of slow when it comes to onboarding data from many different sources in an integrated way. That’s why there’s a data lake in the first place, but there’s still a need for a common interface to the data—especially in terms of compliance, accessibility and security. Logical data warehouses offer a solution to this problem. They let you access all the data from different sources with a common interface and allow onboarding of new data in a fast and easy way. Other than this, we’ll see more data products which will help us enhance the existing data.

We’re going to see more products, such as Azure cognitive services, which will help us implement state-of-the art AI and ML algorithms without the need for in-depth knowledge. Furthermore, many vendors will implement AI in their products to improve the usability of these products. Data scientists’ roles will therefore change, as they become even more like integrators and act as a bridge between departments and vendors.

Many companies I work with tend to do proofs of concepts, and some of them are successful, but ultimately, they fail to put these proofs of concepts into production. Therefore, devops will evolve in this space.

The companies we work with, including Inventx, are very good at defining data defence strategies. We care greatly for our data, as well as our customers’ data. We have many processes in place to protect data. This will change, as companies will also define their strategies to monetise and utilise their data more instead of just protecting data.

We already see more and more of self-service. This trend will develop further. With the rise of Power BI and similar tools, workers will be able to apply simple ML and AI algorithms themselves. Ultimately, co-workers will come to you with ideas and use cases if they know their data.

All in all, I believe that it will be crucial for SMEs within the next two years to develop a sound data strategy to be ready for AI.

Do you agree with my predictions, or do I miss some of them?