The acceleration of digital services has resulted in digital identities being today, more than ever before, a crucial aspect to guarantee secure transactions ecosystems. Fintech companies are those that provide finance services through technology. In the context of the covid-19 pandemic, these companies have increased their relevance and become an agile and scalable solution, providing financial solutions to millions of people around the world.
But how is it that these companies can validate the identity of individuals and small -and medium-sized businesses, who are looking to sign up as clients on their platforms? What mechanisms do they implement to replace the traditional “folders” people used to present at banks and other financial entities, to gain access to their services? How do they guarantee a secure transactions ecosystem?
THE CHALLENGE: VALIDATION OF DIGITAL IDENTITIES TO MERCADO PAGO
Mercado Libre, through their product Mercado Pago, is the leading company in digital and mobile payment services in Latin America. In today’s context, their service has become an ally to thousands of small merchants who, overnight, found themselves in need of a way to offer digital payment solutions to their clients, and upgrade their capacity to sell their services through this company’s marketplace.
According to this study, digital payments boosted during the pandemic in Latin America. In Argentina, for example, digital payments already represent 50% of transactions in small and medium-sized enterprises, and more than 90% of them assure they’ll continue to use e-commerce and e-payments, once the pandemic is over (source: “Argentine SMEs: how they get through the pandemic” (“PyMES Argentinas: cómo atraviesan la pandemia”). In Brazil.Colombia, Costa Rica and the Dominican Republic, 78 percent of the people surveyed stated that they would continue to use digital payment methods even after the pandemic was over (source:Latin America contactless payment).
For this reason, Mercado Pago needs to guarantee a safe experience for its users, while also preventing security flaws, fraud, false identities and other incidents.
Their Know Your Customer (KYC) area requires, therefore, an agile and scalable solution to validate identity, data and documentation for thousands of individuals and small -and medium-sized businesses in the region, with a strong focus on Argentina, Brazil and Mexico, and contribute to the company’s AML (Anti-money laundering) strategies.
THE ANSWER: MACHINE LEARNING TRAINING FOR KYC
MercadoLibre developed its own KYC tool to gain agility and flexibility. This tool uses artificial intelligence models to automate the new users’ onboarding processes, and is complemented by manual validation for those cases presenting ambiguities.
Through Mercado Libre’s KYC tool, Arbusta´s teams add value to this process through a series of key activities: on one hand, they manually resolve tasks the system cannot validate automatically. On the other, they add information and train the artificial intelligence algorithms for their optimization.
Carolina Sangenito Senior Processes Analyst at Mercado Libre, explained that the objective of this project is not only to resolve via human input that which the system cannot validate automatically, but also to generate information to feed its Machine Learning models. “On many occasions, the Arbusta analysts detect improvement opportunities, not only in the KYC system we use, but also in the end-user interface. When we observe that certain information is always completed improperly, we go a little further and improve user experience”, mentioned Carolina.
COLLABORATION AND TEMA-WORK IN TIMES OF COVID-19
Machine Learning Training + Carolina Sangenito team from KYC Mercado Libre at the Arbusta office in Buenos Aires before the pandemic
In march 2020, this project was faced with the challenge of maintaining operations while migrating all it’s teams to remote working modality. The collaboration between Mercado Libre and Arbusta enabled this to happen in an agile manner, while achieving the business´ goals. These months in which both teams, Arbusta’s and our client’s, have been kept in preventive isolation and working from home, the service was not only successfully maintained, but due to the increase in demand, there was an increase in the hours and the number of people dedicated to the project.
With regards to the service provided by Arbusta, Carolina Sangenito expressed: “Arbusta not only offers the service requested; through their innovative orientation, they also challenge their collaborators to think beyond. They stand out for their flexibility, focus on technology and their analytic profiles.” She then added: “Arbusta’s analysts always surprise me with good ideas. Right now, for instance, they were generating a bot to answer each other’s operational questions regarding the project”.
“Arbusta not only offers the service requested; through their innovative orientation, they also challenge their collaborators to think beyond. They stand out for their flexibility, focus on technology and analytic profiles.”
-Carolina Sangenito, Analista Sr. de Procesos en Mercado Libre
At present, a total of 42 analysts in two offices (Buenos Aires and Rosario), 3 QA experts, 3 leaders and a PM, dedicate 5740 hours a month to this project combining human collaboration with artificial intelligence. “Good results come from the predisposition to work as a large team, fluid communication and a shared vision with the client”, expressed Laura Medina, Arbusta’s Project Manager. “At Arbusta we heighten our analysts’ talent development, give them space to unfold their curiosity, to be nourished by the skills of their peers, be productive, and share knowledge. All these efforts make the difference in our service and produce these results which motivate us to keep growing”, added Laura Medina.
Learn more about Arbusta’s Machine Learning Training services.
MORE ABOUT ARBUSTA
> About Arbusta <
> What makes us special <