Artificial Intelligence for everyone
Or How AI + No-Code will transform the work environment
The IT world is continually bombarded with hypes of new technologies and trends that, if on the one hand, prints a frenetic dynamism in this market, on the other promotes the feeling of obsolescence in every professional, driving a tsunami of search for courses and certifications, in an attempt to rush to stay in the market.
I’ve been working in IT for 37 years, I’ve surfed hundreds of trend waves, and I’m going to tell you a secret: all this movement is inflated by the big consulting companies because they need to sell their services to make money. But after Big Data, Small Data, Scrum, Agile, DevOps, Virtualization, Containerization, Cloud computing, Fog, Hybrid Computing, SOA, Micro Services, SaaS, IaaS, PMI, ITIL, and a multitude of new technologies, how to know how to separate what will survive from what is fleeting?
I don’t have an objective answer to this question, but I always follow my instincts and I have almost always been right about how much a new wave comes to stay and stay in the IT world, and with that, I would like to talk about a technology that I have studied extensively for the past 25 years: Artificial Intelligence.
I usually say that I learned to program before learning to speak, and when I look at the world, I see it in code. However, it wasn’t until 2011, after many years of studying AI, that I saw that the algorithms were finally ready to be used in practice. That’s when I started to learn Deep Learning deeply (sorry for the pun) and realized that a revolution was beginning in the software world.
Deep Learning was born to solve a problem intractable by traditional algorithms, which was the processing of unstructured data such as sounds and images, and did this so well that it quickly made all existing algorithms for these applications obsolete. However, what most caught my attention was that these algorithms worked through a mathematical process that, given the program’s inputs and outputs, converged to the solution of the problem.
I’m sorry about being nerdy, let me try to speak more clearly. We are talking about a machine for which I provide an input example, and I say what I want to see in the output, and it programs itself for the solution.
Not everyone is surprised to see this as I was when I understood, but the possibility of creating a program in this way opens ways for anyone to design a program without the need for programming knowledge. Now, to create a program, just assemble a set of input examples with the desired output, and the machine does the rest. For me this is indeed the most revolutionary idea in the software history.
The conception of Eyeflow.AI emerged in 2016 exactly from this idea. A platform for creating Computer Vision applications using AI, which offered a very intuitive and easy-to-use interface, enabling anyone to solve their problems, build their solutions, within their work/knowledge universe, and all without having to study years of programming in arcane codes.
At this time of pandemic, we are experiencing an accelerated growth in the digitization of companies/processes, and the IT world was not prepared for this gigantic growth in demand. We are facing the lack of equipment, lack of chips, and especially the lack of trained professionals in technology. It is in this scenario that this new hype around No-Code platforms arises.
For those who do not know, No-Code are applications that allow a non-programmer to create a “program” to solve their specific demand, in a simple, visual and fast learning way. The idea is very old and we’ve had several implementations of this idea in history with varying degrees of success. An example of this idea that I like to mention is Excel (low-code), where I’ve seen thousands of complex programs (spreadsheets), designed and developed by non-programmers.
As I said, following my intuition and observing trends, I guarantee that No-Code technologies are here to stay, because over the last 20 years we have seen IT become decisive in all professional areas, and this movement is only going to increase. In the future, all people, in any field, will need to be proficient in the use of software and IT tools in general to perform their functions well.
With that in mind, I realized that the scenario was perfect for us to develop a No-Code platform that would enable the creation of Video Analytics applications not by IT people, but experts in their fields of work and business, since using AI programs could be created based on examples/data.
Our goal was to deliver a simple platform where the user can be able to draw what they want, with inputs and outputs flows, and focus on their ultimate goals. From there, it feeds the system with a set of examples and the machine programs itself for the solution. Afterwards, the user observes the results and when he finds an error, he corrects it and feeds back the machine, which in turn continues the learning process until it finally gets it right.
Eyeflow.AI was born to democratize the use of AI, make it available to everyone, and help people around the world to create their Video Analytics solutions. We are making this technology available so that anyone or any company can create, test, improve, publish and sell their own Computer Vision applications.
The main application of this technology today is to automate the repetitive and tiresome tasks of visual monitoring. We have cases in industrial inspection, safety monitoring, and automation of various tasks that depend on vision. I confess that every day I am still impressed to see how each person watches an EyeFlow presentation and immediately thinks of a new application for a specific need in a new market.
And to those who fear that machines will take their jobs, I always answer that human beings, with their extraordinary brains, were born to create solutions to problems, and machines are those creations that will do the job. Think, for example, of tractors in the 1950s, and how technological evolution has made our lives better over the last century.
