• With artificial intelligence today we address needs that cannot be neglected anymore. We do not replace anything: we integrate into the existing economy to achieve economies of scale in 10, 12 months. Those who do our work, addressing specific needs on manufacturing and media companies, have a great role at this time.
To speak with us so is the new CEO of Ammagamma, David Bevilacqua, old knowledge of the national IT world. The experience he had as the director of the South Europe region and as the head of Cisco, culminates in a managerial path followed by the creation of Yoroi (cybersecurity company now in the Tinexta group) and the activities
Then there was the meeting with Fabio Ferrari founder (in 2013) of Ammagamma, when still the Modena company was called Energy Way, who co-opted him first as a board member, then entrusting him the role of general manager
The narration of artificial intelligence
Enough with the use of metaphors of humanoids and androids: Â this company tells artificial intelligence for what it is: applied mathematics Â Â.
Today, according to Bevilacqua, to bring artificial intelligence into production, we need to reread the physical and transactional processes to enable what people do normally.
• We make mathematical models by revisiting customer processes according to the objectives that customers give us We use their data, add new data and efficiently work • • We use their data.
Ammagamma is doing this with about sixty people, we said, among production figures, data scientist, data architect, the commercial structure and staff. A typical size for a scale up that grows organically, but that has customers that are yes large companies (Telecom and Eni: the presence in the utility world is confirmed by the fact that before Ammagamma was called Energy Way) but also many small and medium companies
They are manufacturing and industrial companies, he explains, that need to find new forms of efficiency and optimization.
From data lake to mathematical model
For these medium Italian production companies, what does Ammagamma do?
First he understands the goal they have, explains Bevilacqua. ›We have 70 use houses, with which we understand the customer’s goal and compare it with the available data: this is the first point to be addressed. From here we build a consistent data lake. Then we add the exogenous data that do not depend on the company but that cling to the business model. For example, the weather. After that, we write the mathematical model, we turn it into software and put it on devices: a plc, a server a cloud
This model is then from Ammagamma maintained, trained, upgraded, integrated, evolved.
Examples of Ammagamma’s use case are predictive controllers for purification plants, image analysis for quality control, data analysis for energy management, optimized production planning, mathematical models for multisite, robotization
In Modena the experimental school of artificial intelligence Ammagamma and LM3 Mattarella of Modena gave birth to Lucy (the name comes from the Australian language with intelligence), the first Italian experimental school of artificial intelligence for The aim is to educate young people and adults to understand and use data and artificial intelligence technologies, in a conscious and responsible way with experiential experimental activities. The project will allow students and teachers to discover the potential of technologies not with an extra activity but with a curriculum learning path. Students will experience theory in practice by rediscovering the value of the sensory dimension in the learning experience. For Pietro Monari, head of the Education projects of Ammagamma today , needs a multidisciplinary approach to understand the artificial integrity. This is the ambition of our school and the educational vision of Ammagamma.
System integrator of themselves
The business model at the beginning was of a classic consultential model, based on the proof of concept. ♪ Now you create code, we apply a model from software house, with system integration. We are also system integrator of ourselves, when for reasons of dimensionality the customer company cannot have it.
Ammagamma performs software activities to create models of artificial intelligence with developers, UX designers, integrators, following a model that Bevilacqua calls \’modify the shelf Â Â: from a common base you switch to customization.
♪Now we have nine encoded solutions (inventory, predictive control) that we replicate depending on the customization model. Solutions for most horizontal and in some cases vertical for industry. Replicable solutions
Ammagamma goes where there is no data scientist
To large enterprises you don’t have to explain what artificial intelligence is, they already know, claims Bevilacqua.
• Companies that are not equipped with data scientists need to be approached by Ammagamma, with a vocabulary that is understandable. To do this today we have a commercial network with a double approach, with large customers and people on the territories, with geographical responsibilities.