EAE News
Ángel Galán: “Big Data applies to all the business units in a company”
13 de December de 2019
Big Data is clearly a big factor in the present and future of all organizations. Using this data well will determine whether a company climbs and embraces greater business opportunities.
Based on this premise, EAE Business School organized a webinar entitled ‘Big Data for Decision-Making’, led by Angel Galán the Director of Big Data & Artificial Intelligence at Kabel.
Before joining Kabel, Galán was the Director of Business Intelligence and Data Analytics at Correos, and he joined the company by coincidence. “They were in a transition phase towards a new analytical model. They had some really specific needs and wanted to develop this area”.
Since entering the world of Big Data, the most significant challenge has been keeping up to date. “Technologies evolve extremely quickly. The world of data is increasingly broad, and it is hard to maintain a sufficiently comprehensive vision to give guidance in corporate and global strategies”.
Ángel Galán began the webinar with the idea that companies find it hard to offer individualized products and service, and data in itself does not imply a good decision. “It is the interpretation of the data and how it is applied in the business process that generates value”. Moreover, he emphasized that, although we deal with a huge amount of information nowadays, the data that we use in our analyses remains constant, and most companies only use 1% of the information.
Evolution of the world of data
90% of the data available in the world has been produced in the last two years. “We create a huge quantity of information. We have a large number of elements that we interact with and, most importantly, the amount of systems that are no longer just personal, but rather systems such as the IoT, which generate a massive quantity”. To illustrate, he used the example of the human genome project. “Its objective was to identify and map the approximately 20,000-25,000 genes of the human genome from a physical and functional perspective”. According to the professional, multiple teams were needed, around the world, working in coordination. It took about 10 years to map the entire human genome, at a cost of 1 trillion dollars. “In contrast, in 2015, it took us a day and cost 100 dollars. By 2025, we are expected to be able to do it within a second at the cost of a single dollar”. As such, the increase in the variability of information processed in various volumes has led to an absolute democratization that has reduced costs and “eliminated barriers to access, both in terms of data generation and consumption”.
He also gave an overview of the evolution of computing capacity. “The capacity in 2001 was equivalent to the brain of a mosquito. In 2010, it was equivalent to a small mouse. By 2020, it is expected that, for 1,000 dollars, you will be able to buy a computing capacity equivalent to the human brain. By 2050, we will be able to buy the capacity of all human brains put together”.
In Ángel Galán’s opinion, one of the essential features is the importance placed on the value of information depending on time. Information is more valuable the closer it is to its creation. “To a large extent, the volume of data is linked to the diversity of types of information that we can have nowadays”. He also emphasized the importance of being able to work with all this data in a structured form, because this enables to apply this algorithm with our own images on unstructured data. “This completely changes our philosophy”, he explained. In the opinion of the Director of Big Data & Artificial Intelligence at Kabel, Big Data, which originated as a technological solution, is now applied to all the business units in a company. Moreover, “we increasingly see mixed profiles within these business units, which enable us to work with data in a far more flexible way”.
Big Data & Big Problem
Angel Galán acknowledged that, in our day-to-day working lives, we face difficulties on our projects because these high expectations are often not met. The reality is that 60% of Big Data projects do not fulfil expectations, explained Gartner. “44% of Big Data projects are cancelled and the expectations of breaking even on the ROI of 3.5% is 0.5%”.
In the professional’s opinion, this happens because there has been a sharp increase and trend to overvalue a technology initially. “We need to achieve a certain level of internal maturity within the company in order to be productive”. This is precisely what happens in any type of initial phase with respect to a technology: “In this kind of situation, we have to be realistic. Generating a predictable return is not simply a matter of implementing the technology. Sometimes there are changes in processes and a lack of maturity to consider when incorporating the technology into departments”. Therefore, he concluded that Big Data is here to helps us and we need to increase the value of this information.
Analysis in decision-making
Ángel Galán drew a distinction between five types of analyses: descriptive, diagnostic, predictive, prescriptive and, lastly, artificial intelligence. Firstly, we have to deal with the raw data, which is as yet unprocessed. “Diagnostic analysis is an advanced analytical approach that examines data or content in order to answer the question “What happened?”.
Predictive analysis is an advanced analytical method that examines data or content in order to answer the question “What is going to happen?” or “What will most likely happen?”. This analysis is characterized by the use of techniques such as regression models, multivariate statistics, pattern identification, and predictive and prognostic models.
Prescriptive analysis is an advanced analytical approach that examines data or content in order to answer the question “What should we do?” or “What can we do to make it happen?”. It is characterized by techniques such as graphic analysis, simulation, complex event processing, neural networks, recommendation engines and automatic learning.
“When we bring all these factors together, this is when artificial intelligence comes into play, which is a factor that differs from prescription because it is far more actionable”. When we speak about artificial intelligence, the Big Data expert explained that the image that comes into our head is not an algorithm, but rather some type of robotic future. “In reality, we don’t see it simply as a matter of automating processes. It is science that enables us to train systems to emulate tasked carried out by people, through a process of learning and automation”.
Watch the full video at the following LINK.