SIDIS accelerates data-driven futures with MongoDB

THEIR CHALLENGE

Using MongoDB to handle data proliferation and deliver flexibility

Launched in 2019, SIDIS is a young, agile, ambitious business based in Porto, Portugal. It specializes in providing a universal data management platform that unifies and enriches business data and transforms it into the real-time operational insights that clients need to make informed commercial and strategic decisions.

“Everybody talks about how exciting and powerful AI is, but if the data foundation doesn’t work, everything falls apart,” said Carlos Marquez, Director of Growth at SIDIS. “That’s exactly where we position ourselves. We centralize different data sources into our platform, create algorithms that cross-reference them, and then put the results into action.”

Use cases can range from the ability to hyper-personalize communications, to insights into customer behaviors and buying patterns. SIDIS’s customers range from Migros, Switzerland’s largest retailer, to one of the largest banks in Latin America.

“They’re large corporations with capable people, but if they wanted to develop a solution similar to ours it could take anything up to five years, by which time the technology is already legacy,” Marquez said. “We specialize in taking all that structured and unstructured data and using our database to do all the heavy lifting behind the scenes.”

It is a service that is both foundational and future-facing. As businesses worldwide seek to embrace high-power analytics and AI, SIDIS is placing itself at the forefront of enabling change and opportunity.

SIDIS launched using MySQL; however, the business and the volumes of client data it was handling grew fast, and it soon became clear that the company needed a more powerful and adaptable option.

“We were managing around 1 million daily transactions from just 20 sources, compared to 80 million daily transactions from over 500 sources now. We saw that we weren’t capable of handling everything ourselves, and we also needed greater flexibility to meet different customer requests,” Marquez added.

OUR SOLUTION

MongoDB enables implementation of a hybrid on-premises and cloud-based solution

SIDIS already had developers with MongoDB experience. While that played a key role in the decision to implement it, Marquez pointed out that there were additional important factors.

“We could take advantage of our developers’ knowledge to make the transition fast and smooth,” he explained. ”Also, we needed a solution that could work both on-premises as well as in the cloud. That’s a unique selling point that MongoDB’s competitors just aren’t bringing.”

SIDIS saw that MongoDB Atlas would work excellently for many clients, however the compliance requirements of key government and banking clients required data to remain in premises and within the country. The option to operate either with MongoDB Atlas or on-premises was, therefore, critical.

“Migros is a little different in that it’s a mix of its own cloud, our cloud and on-premises,” Marquez said. “MongoDB enabled us to implement that hybrid structure very quickly while ensuring data security and encryption. That was very important in gaining Migros’ trust; in fact, I’m sure that we wouldn’t have won the business without it.”

Importantly, SIDIS was able to implement MongoDB and incorporate its capabilities immediately without having to expand or reskill its existing internal teams.

“We started working with both structured and unstructured data, which we thought might bring challenges,” Marquez said. “But there is such a wide knowledge base within the MongoDB community that it was easy to outsource any tasks that we couldn’t handle internally.”

Another key requirement was meeting seasonal and peak scalability spikes in demand. “In the banking industry, transactions tend to be focused around the end of the month when people get paid, while other weeks don’t need the same resources,” Marquez noted. “We knew that the ability to adapt and scale capacity on a day-to-day basis would be very helpful to us.”

OUTCOME

Managing high volumes of data effectively opens up new opportunities

With MongoDB as its foundation, SIDIS is now enabling a leading bank in Latin America to access and process data from a range of sources, including its legacy banking systems.

“The bank had been using spreadsheets to build reports,” Marquez said. “It could access the important core systems, but it was nowhere near a complete overview of the business and its customers. With SIDIS and MongoDB, the bank can now access and utilize 180 different data sources and has developed more than 1,200 indicators, categorized into traditional statistical models and AI-enhanced metrics. Traditional indicators are based on fundamental mathematical and statistical methods, while AI-powered metrics leverage machine learning algorithms and advanced analytics to provide deeper predictive insights.”

The bank can now send up to 30 million personalized emails each month to its customer base. Designing, processing and executing campaigns now takes minutes instead of days, and to date, the bank has processed 50 billion data points and over 280 million emails. Processing and reporting on the bank’s 80 million daily transactions also previously took several days; with MongoDB, insights that inform key decisions and improve operational efficiency are now available in real time.

“One key report that the bank uses was previously limited to around 8,000 of its largest business accounts,” Marquez added. “Now that report covers 18 million accounts, and instead of only being produced monthly it’s available at any time in seconds. That gives the bank a level of data transparency that it previously just didn’t know was possible.”

Working with SIDIS, the bank has now developed an improved credit scoring system that uses AI to assess customer backgrounds and more accurately predict their behavior. It also developed algorithms to recommend product offerings based on customer behaviors, such as website navigation and email opening rates.

“We’re also enabling predictions about product performance, quality and likeability, and carrying out sentiment analysis to evaluate the success of marketing campaigns,” Marques explained. “We’ve even included a generative AI model that translates the content of each dashboard and table into explanations in natural language.”

Unsurprisingly, SIDIS is expanding—and expanding fast. It expects to add over 1 billion transactions over the next three years as it onboards around 20 new banks, and has added North America to its main targets of Europe and Latin America. The company has also received recognition from Marcelo Rebelo de Sousa, the President of Portugal, for its pioneering work in AI.

“We’re growing in banking and insurance, as well as the telecoms, retail and automotive sectors,” Marquez concluded. “And the support and partnership we get from MongoDB is something we put at the forefront of our offering.”

To learn more, visit MongoDB Atlas.