It is expected that this year alone, companies will invest more than 500 billion dollars for the implementation of artificial intelligence solutions in their digital infrastructure. Moreover, the study conducted by IDC in 2002 states that this amount will increase by 20% every year thereafter.

However, many companies that have already followed suit are slow to see results commensurate with the amounts and energies spent and give bad publicity to AI, which is often perceived as a magic potion.

In my opinion, the first goal of implementing AI should be to allow managers to speed up their decision-making, while reducing their risk of error. For this to happen, it is important that the created system has full access to all the data it needs. However, in reality, for various reasons (including confidentiality), the accessible data is only partial. Also, they may not reflect future data. In short, the actual availability of relevant data compromises the validity of the exercise.

Second, it is imperative that data experts collaborate with IT infrastructure and operations experts. There must be a synergy so that the decisions that arise from the operations of the AI are applied. Too often, the best innovation projects do not happen, simply because they were those of a few, rather than everyone in the company. Generally, people are busy and preoccupied with their daily workload. They apply to work whose results are measured overnight, even hour by hour, while the projects resulting from AI are sometimes perceived as less tangible, because they are spread over time, and therefore not urgent to deal with.

Third, it is important that the objectives pursued be clear and based on real problems to be solved. Many companies have somewhat vague expectations and the results are to match.

To better explain myself, I will use an example. Let’s imagine that your stocks are always too high and you know that your competitors are doing better than you in this regard. Also, you decide to use artificial intelligence to manage your supply chain and you are targeting, within 12 months, a 30% drop in your inventory*. With such a focus on a measurable objective, and if you know how to apply the right actions correctly, I’m going to bet that you will be satisfied with the return on your investment, as long as the solutions have been correctly implemented.

It is true that AI is synonymous with complexity. That said, it is understandable that many companies continue to hesitate before starting such a project. Yet there are relatively easy ways to redesign your digital infrastructure to be an optimal environment, designed for AI. This could be an issue for the competitiveness, or even the sustainability of your business.

If you hesitate despite everything, do not hesitate to contact me, it will be a pleasure to discuss with you on this subject, or on any other of an IT nature constituting an issue for your company.