English
Español Today, it has been held the conference-symposium on Artificial Intelligence applied to business, moderated by Professor José Luis García del Pueyo at the San Telmo headquarters in Seville. For the occasion, participants in the talk were Carlos Parra (Emergya Cloud Director), Javier Pérez Trigo (Iberia Partners Team Lead, Google Cloud at Google) and Alejandro Gómez Morón (Emergya Principal Engineer).
First of all, the Iberia Partners Team Lead, Google Cloud contextualized the new digital environment. He explained that while interest in Big Data has been sustained over time, searches around Machine Learning have increased over the past two years. He also noticed how Artificial Intelligence not only repeats patterns, but also learns from models and data to accomplish creative solutions never applied before to specific problems. Machine Learning brings also a radical change in the way we code, the rules to be followed are not specified, but guidelines for the machine to learn. Javier Pérez Trigo mentioned a revealing fact: "90 % of the data in any organization is unstructured. Machine Learning can help you get the most out of it."
On the other hand, Alejandro Gómez Morón from Emergya showed how the Machine Learning model of Emergya can help in business. For example, how real-time detection allows stores to customized the shopping experience; for that, he analyzed a photograph to quickly segment the audience, which leads to an increase in the average ticket purchased. Other applications of Machine Learning that he displayed were: the validation of official documents such as ID documents to fight fraud; analysis and classification of texts; or analysis of videos in real-time, very useful for security companies, since it allows to identify false positives and to activate when there is a real risk, among others.
The last turn to speak was for Carlos Parra, Account Manager of Emergya. First, he talked about the key to implement Machine Learning. In this sense, he said that the essential is that there are enough data to generate models. For that, he pointed out three ways:
- Existing pre-trained models
- Customization of pre-trained models
- Developing own models
He also set out four questions that every business should answer before implementing Machine Learning:
- Are you recording and storing the relevant information?
- How can you use that information for competitive advantage?
- What data can you get from other platforms?
- How to apply a Machine Learning model step by step?
Following the speaker’s turns, Professor José Luis García del Pueyo opened the floor for questions, encouraging an interesting and enriching debate
