If you are reading this post, I encourage you to start your AI or ML project as soon as possible. As my partner would say @Jbu3s0 “If it is done in Power Point it is Artificial Intelligence and if it is done in Python it is Machine Learning”.
Stop spinning, the road is made walking and, if you are like me in a large corporation, you will find a thousand obstacles but eventually you will arrive. In my case, I decided to start this road in 2015 with 4 other entrepreneurs. Our story comes from a few years ago, but to not entertain much, I leave here a section if you want to take a look 😉
Even so, I feel bold enough to put myself in your place and summarize some pebbles that I found along the way, and that, almost certainly, you will also find. Preventive computer is worth two, so here I leave posters indicating where the gravel.
I start the reverse of most tips on how to start IoT in a corporation through a startup. First, beware of technology, what generalists treat as a commodity and leave it at the end of the lists of where to pay attention, to me it seems a capital point. The basic concepts of the AI, are more than 60 years old and will pass many others until this technology is intermixed in a fluid way in our day to day.
This means that a lot of smoke is sold, and some church arches easy to make and which are not. Now, selecting appropriate platforms from the beginning, you will find great potential hidden in your data and the business processes that combine them to create really new value. So, immerse yourself in the current technology and its short roadmaps and, above all, in the medium term, this took us 1 year. Choose well in relation to what you want to create, change then it will cost a lot.
When attacking a “pain” in your corporation, choose one really anguished, anchored and expensive to solve and once there do not allow you to be distracted from it, in large corporations it is easy to end up shooting everything that moves. The solutions that require AI and IoT, the more they are used, the more value they create, and of course the other way round. So the only way to give value is to identify very well where that intense “pain” is.
In our startup, we do not understand another way to roll up our sleeves other than Agile, Scrum and Devops. Actually we did not want, however, we had no choice. When we need to associate our idea with outside talent, other startups … the people who disembarked in our offices only spoke this language.
So, as “force hangs”, after 3 years working with concepts such as lean or agile, do not be scared!
I have learned that these methodologies are nothing more than putting a lot of common sense on how to tackle a complex technological challenge and mitigate the “pain” of a need.
We must not forget about people. Look at this Gartner data for 2022, that is, right now:
- 1 out of every 5 workers who are mainly dedicated to non-routine tasks will depend on processes based on Artificial Intelligence to carry out their work.
- 2.3 million jobs will be created and 2,800 million hours will be saved worldwide.
- The energy utilities sector will benefit from the AI without suffering a net annual loss of jobs. From the point of view of the value that is created, it will come from gains in efficiency and income opportunities as part of new business models driven by the knowledge of the data.
And this is what my last and perhaps most important advice is based on:
It is the one related to ourselves, the people. On the one hand, those who have the pain on their side and who are in the business of changing, not only the way to solve it, which is easy, but to change the way they work from tomorrow. Something much harder to do.