[MUSIC] Today we're talking with Paola Cagliani, Head of Big Data and Advanced Analytics at Eni Gas and Power. The Italian name is Eni Gas e luce, one of the most innovative energy providers in Italy. Paola, thank you so much for being with us today. >> Thank you for inviting me. >> Eni Gas and Power was initially part of Eni, the biggest energy provider in Italy, that has been formally established as a separate company in 2017, to keep up with rapid changes in the energy sector. Paola, can you please give us a brief overview of what the company does? >> Yes, Eni Gas e luce is a energy retail company. The purpose of Eni gas e luce is to sell gas and power to our customers that are retail and business customers. Eni gas e luce operates in four European country with 1,600 employee. In Italy, the headquarter is here in Milano, we have about 8 million customers and the mission of the company is to help customers to use energy better and less. >> With so many customers, you must have a huge amount of data. Can you tell us about the kind of data that you have? >> Yeah, as a everyday company in an energy world we have data about contact, about consumption of gas and power, about payments. But those about interaction that the consumer has with us on the different channels. Could be web, could be stores, or call centers. In term of usage of this data, we use this data for the manage our web business, so to run our business, to analyze the performance on some particular KPI, like financial KPI, marketing or commercial KPI, operational KPI. Then we use the data to identify new trends or new business opportunities. And those are to create new products and services. I think that Eni gas e luce is a data-driven company. So this is why we decide to give access to the data to hold the employee data need it. And we invest a lot in data visualization and reporting tool. >> Have you changed the way you collect and use the data in the last few years? >> Yes, I think that we've completely changed the way we collect and we analyze the data. In 2015, or the year before, we started analyzing data introducing data visualization and reporting tool. Immediately afterwards we started business intelligence technology in order to do better analysis to predict some phenomenon. In the last year, in 2018, we decided, this company, Eni gas e luce, decide to start a new program called big data advanced analytics with the aim of introducing the big data infrastructure like the technology on which develop knew advanced analytics models. The ambition of this program is first of all to combine together a structure data like a data coming from relational database, CRM or billing system. With unstructured data, like data coming from social, or coming from Internet of things, or coming from images. The second ambition is the introduction in the company of machine learning, and in particular, deep learning technique in order to apply this technique to develop a better predictive model. And for the reach of this ambition, Eni gas e luce is investing in developing data scientist, data engineer and the data analyst profiles. >> Aside from these data analysis capabilities, do you have a framework in mind when you analyze these data? >> Yes, usually we approach the analysis with a specific goal in mind. Usually, it often happens that business units have a question to which they are not able to give a precise answer. We analyze all the data available and we try to find the answer to those questions. For doing it we apply simple, sometimes simple analysis like descriptive or diagnostic analysis. Sometimes we apply complex analysis like predictive or prescriptive analysis. On advanced analytics we approach it with what I call use case driven approach. Which means that we identify the opportunity to generate value applying, advance analytics models. In general the use case are use cases that they cover all of the areas of the business. So we have the use case that covers the needs of marketing and sales, needs of operation but also needs of information technology. So to answer your question: "it's yes", we have a framework that helps us to identify to understand better the business need and to apply statistical machine learning or deep learning technique to reach the goal and to generate value. Sometimes happens that analyzing the data we discover that there are some unexpected patterns. And this for us a big opportunity to go deeper on this unexpected pattern and do drill down analysis. >> This is very interesting. Can you give us an example of something that you discover from the data that you were not expecting? >> As I told you before, when we combine data coming from different sources sometime we discover some unexpected relationship that surprise us a bit. And recently, for example, the team running analysis on online customer. That are the customers, who sign a contract with us using the online properties. And analyzing, we realized that they have different characteristics than the customers that sign a contract with us using a non-online properties. In particular, we discovered three characteristics that differentiated the customer. First, that the online customer consumption of consumer power is higher compared to the rest of the customer base. Second, that they are more precise in terms of payment. And third, they are more loyal in terms of generate. With this analysis, we were able to decide better our immediate investment in order to attract the customer with similar characteristics. >> That was a great example thank you so much for sharing your experience and insights with us Paula. >> Thank you for inviting me.
