GETTING CLOUD AND ML
INTO THE DNA OF NEXI

FROM WORDS TO FACTS

Despite proof and evidence of the benefits of cloud migration, skepticism remains about its security, slowing the process of digital transformation in safety-critical contexts such as Banking, Digital Payments and all those companies that have stringent regulatory requirements. Inspired by its customer experience, Data Reply enabled Data Analysis and Machine Learning on AWS Cloud for Nexi, the largest Italian PayTech Company, bringing quantity and quality of data and leveraging Artificial intelligence-based technologies, resulting in major impacts in customer’s capabilities in areas like Fraud, Risk Management, Marketing and Operations, in a safe and compliant way.

The project provides value to customer’s business through all the benefits of a resilient, cheap and scalable Cloud solution on Amazon Web Services, still security-compliant.

NEXI, INNOVATION IN DIGITAL PAYMENTS

Nexi works with Banks to provide solutions for making and accepting digital payments that are innovative, simple and secure for Individuals, Business and Public Administration. Nexi manages about 41,3millions of payment cards and 5,5 billions of transactions per year, counts 890k merchants, 1,4 millions of POS terminals, 446 billion euros in transactions, 13,4k ATM in Italy and 420k companies members of Corporate Banking services (Nexi - Investor Relations ).

Therefore, Nexi is a customer intrinsically oriented to innovation, investing in technology and owning large amounts of data, from which useful information can be extract, enhancing business knowledge and building better customer experiences, improving operational efficiencies, delivering new products and finally driving successful actions through informed decisions.

Nevertheless, digital payments and banking businesses have critical issues on data safety and security, slowing down cloud transformation process and migration to the cloud.

In need of a flexible and scalable solution

The huge amount of data managed by Nexi is ingested from several sources on its systems and resides on on-premises facilities, making information scattered and hard to access and consult. The challenge lies in data distribution on heterogeneous source systems, whose access is regulated by strict laws concerning security and privacy, but also on a great workload variability. All this leads to a cloud solution, to leverage flexibility and scalability of provided services.

The final target is creating a complete and centralized analytical infrastructure for data analysis and exploration, to identify useful information, create and industrialize machine learning models suitable to extract this information, all on a single platform, thus creating a big datalab for data scientists and analysts, accessible and scalable. The architectural solution has been designed together with Nexi Data & Analytics departments and fully developed by Data Reply, accordingly to client’s guidelines. Sensible data are processed and encrypted on-premises, thus making storage of data on cloud provider safe and compliant with GDPR laws.

BUILDING AT SCALE WITH AWS CLOUD

Amazon Web Services is the chosen provider, as the client needed a scalable and serverless computing model to run ML models without worrying about infrastructure maintenance, still maintaining flexibility and cost-effectiveness and AWS provides a wide range of different cloud services which cover all critical aspects and requirements of the project.
These services are tightly integrated and provided a convenient way to build the required solution flexibly and powerfully. This cloud solution is low maintenance and reliable, accessible and with almost unlimited storage space, but above all it’s scalable, handling a growing amount of data seamlessly. Larger and larger amounts of data and processes are brought to the cloud, thus identifying new use cases for the company, discovering new lines of business, improving marketing and selling strategies and finally enabling effective decision making to obtain best business outcomes.
Today 20TB of data have been already ingested and 300GB are ingested daily on the cloud. Tens of millions transactions are analysed every day at scale. More than 10 machine learning models have been developed, among them block prevention, churn detection and anti-money laundering. Data Reply is currently supporting client’s data science team in data analysis and ML modelling.

REPLY VALUE

“In our cloud transformation process, we were looking for a partner with a strong experience on data analytics on cloud and a proven capability to deliver solutions in a production environment. We recognized in Data Reply a company capable to support us in this process, with a profound knowledge of the technological aspects of our cloud platform of choice.”

Stefano Gatti, Head of Data & Analtycs at Nexi



“What we found has been a team of highly skilled professionals with a positive attitude, as well as a broad experience on dealing with different data science problems from a methodological point of view. All these aspects have been fundamental for us in order to reach our goals.”

Alberto Danese, Head of Data Science at Nexi


  • strip-0

    NEXI S.p.A.

    Nexi S.p.A. is an Italian company with international profile offering digital payments services and infrastructures to banks, companies, institutions and public administration. Nexi’s mission is to change the way Individuals and Businesses make their day-to-day payments and collections, making all payments digital, easier, faster and more secure.

  • DATA REPLY

    Data Reply is the Reply group company offering a broad range of advanced analytics and AI-powered data services. We operate across different industries and business functions, enabling them to achieve meaningful outcomes through effective use of data. We have strong competences in Big Data Engineering, Data Science and IPA; we build Big Data platforms and implement ML and AI models in a manner that is repeatable, efficient, scalable, simple and yet secure. We supports companies in combinatorial optimization processes with Quantum Computing techniques that enable an engine with high computational performances.

    strip-1