Colombians Esteban Marín, Eduin H. Serna and machine generates inferences and enriches the Andrés Hernández founded DILE in September experience of clients, with no intervention by 2017, a company providing technology with a the human being in the process. platform offering inancial solutions through predictive analysis of large non-structured Beyond a business intelligence data transforma- databases. tion development, DILE contributes to improve DILE eficiencies generating a Learning Capital. Following a B2B business model, the data entered www.dile.co by a bank to DILE is processed to produce the Regarding clients, DILE has two types of approach. indings in the reports, providing predictions. No On the one hand, it has the companies already individual intervenes; therefore, the client has with a speciic need in their market; on the oth- to worry only about the information it feeds. In er hand, it works with the companies willing exchange for a monthly fee, access is given to to approach technology, but they do not know the platform, solving this process through algo -how technology can help them. Eduin H. Serna, rithms. This allows the enterprises/companies co-founder and CTO of DILE, explains that once afiliated to DILE to know their clients and predict they do the implementation and a hypothetical their behavior to offer them speciic services. To pilot with outcomes, the client gives real data and develop in the future and prove its technology, the pilot is launched to know how effective the this FinTech also has pilot programs in other tool is. This stage is called “training” and eficiency sectors, such as retail, health, transportation of the company is not only in the results of its and telecommunications. AI, but in the delivery time, since, for example, while the inancial institution performs a similar The idea of creating this company emerged after action within six to twelve months, DILE does detecting the need to transform the inancial it within weeks. industry processes, left behind in technological innovation for the last ifty years. It appears from Although startups in the FinTech industry have this that DILE main innovation is the imple-the capacity to raise million dollars in risk capital mentation of the deep machine learning (DML) funds, DILE has tried to inance itself by itself, engine, AI based on reinforcement learning with with only a irst investment of 3,000 dollars at the more than 700 machine learning (ML) models beginning. DILE currently operates with positive and evolutionary strategy, constantly improving lows and still inances technology development itself this way on the processed information. This from the income generated by its sales. is linked to another basic ML concept, known as “unsupervised learning”, this takes place when DILE is the way in Latin America to change the the machine is no longer taught, but the machine inancial industry through artiicial intelligence. learns by itself and, based on such learning the A TWO-WAY PATH87