[Hands-On Workshop] How to Build Streaming Agents with Flink, Claude LLM & Anthropic’s MCP | Register Now

Stone Pagamentos Logo

Stone Pagamentos Takes Fintech From Patchwork to Platform

Accelerated integration timelines by 40-50%

Confluent reduced provisioning to minutes, enabling deployment of over 100 connectors

Empowered Stone to move quickly in an increasingly competitive fintech landscape

"Confluent makes everything easier. Being able to handle all the cluster management through one platform is a game-changer for our team—it frees us to focus on building value, not just managing tools."

Allan Oliveira

Site Reliability Engineering Manager, Stone Pagamentos

Shoring Up Stone’s Real-Time Data Streaming Infrastructure

As one of Brazil’s leading fintech companies, Stone Pagamentos offers comprehensive solutions, like point-of-sale (POS), banking, and credit payments—everything a business needs to sell, manage, and operate. From small merchants to large enterprises, Stone serves a wide range of businesses with a strong focus on technology, innovation, and customer service. 

Stone’s horizontal platform team provides a robust, scalable, and reliable messaging solution for its engineering organization, empowering over 1,500 developers with the tools and infrastructure to build real-time, event-driven applications. The team is the backbone for Stone’s real-time data infrastructure, directly enabling the company’s core revenue streams. Whether it’s physical card transactions, virtual payments, or PIX transfers, the platform team supports critical business functions. “We're the engine that allows Stone to move quickly in a competitive fintech landscape, while maintaining the reliability that financial services demand," says Allan Oliveira, Site Reliability Engineering Manager. 

Despite the platform team’s vital role in the organization, Stone's fragmented streaming landscape hindered its ability to respond swiftly to business needs.

Navigating Stone’s Data Maze to Unlock Fintech Agility

Stone embarked on its event streaming journey several years ago with a variety of platforms. However, by 2023, it became apparent that the company needed a more standardized, scalable, and cloud-agnostic solution for managing real-time data flows across its diverse array of systems. Thus began Stone’s partnership with Confluent.

Before Confluent, Stone's streaming landscape was a complex patchwork of solutions, including RabbitMQ, AWS SQS, Azure Service Bus, and MSK, often based on a team’s specific needs or preferred cloud providers. While this approach offered short-term flexibility, it ultimately led to significant operational challenges that hindered the company’s progress.

"The real problem wasn't just the technical complexity," explains Oliveira. "Each technology had its own monitoring, alerting, and deployment processes, so our infrastructure team was constantly context-switching between different tools and troubleshooting methods. Knowledge wasn't transferable and onboarding new developers took longer. "

More critically, this fragmentation severely impacted the platform team’s ability to respond to business demands quickly and efficiently. Incidents often required multiple specialists to trace data flow across disparate systems. Scaling individual components demanded different approaches and timelines, transforming capacity planning into a recurring nightmare. Furthermore, cross-team integration projects became tedious negotiations between conflicting messaging paradigms, rather than straightforward technical implementations.

"Essentially, we were spending more time managing our tools than actually building value for the business," Oliveira elaborates. "In fintech where speed and reliability are everything, this was becoming a competitive disadvantage."

To address these critical challenges, Stone identified key requirements for its new streaming solution:

  • First, the company needed a unified platform that could consolidate its messaging infrastructure without vendor lock-in. Given Stone's multi-cloud strategy, any solution had to work seamlessly across AWS, Azure, and other providers as the firm scaled.

  • Second, Stone needed enterprise-grade operational support—managed infrastructure with proactive monitoring, scaling capabilities, and built-in security. The hidden costs of its fragmented approach were becoming unsustainable, and Stone needed its teams focused on building features, not babysitting middleware.

  • Third, the company  needed governance capabilities that provided visibility and control without slowing development. This meant centralized monitoring, standardized access controls, and company-wide policies while still allowing teams to move fast.

Stone’s legacy messaging systems also imposed significant operational burdens due to its inability to dynamically handle fluctuating workloads. During high-demand events, such as sales peaks or seasonal surges in transaction volumes, the company’s systems struggled to maintain performance and stability, often requiring frequent manual interventions from the platform team.

How Confluent Ignited Stone’s Real-Time Data Revolution

Stone found its answer in Confluent Cloud, the fully managed, cloud-native architecture that directly addressed its critical needs for scalability, high availability, and streamlined operations. Confluent enabled the company to move away from constant manual interventions during peak loads, and its predictable pricing model helped with cost management, allowing for better use of resources while maintaining a robust and reliable messaging environment. 

"Confluent makes everything easier. Spinning up resources is straightforward, the auto-scaling works really well, and being able to handle all the cluster management through the Confluent interface is a game-changer for our team," shares Oliveira. "What really sold me on the managed streaming approach is how stable and dependable the platform proved to be. We moved some pretty sensitive data and mission-critical apps from our old on-prem setup to Confluent because we trusted it that much. The Confluent data streaming platform represents where the industry needs to go—taking something as powerful as Kafka, but removing all the headaches of running it yourself. It's one of those solutions that just makes sense."

In its banking operations, Stone had previously encountered significant challenges with the availability and reliability of its legacy messaging tools, which frequently contributed to operational incidents. These issues exposed critical gaps in its infrastructure, particularly concerning the seamless, real-time data processing and communication vital in the banking sector. 

Confluent solved these pain points with proven high availability, robust scalability, and advanced features tailored for enterprise-grade messaging. This transition enabled Stone to establish a more resilient and efficient messaging environment, enhancing its ability to meet the demanding requirements of its banking operations.

Two Confluent Cloud components have been particularly instrumental in Stone's modernization journey:

  • Confluent Connectors: Stone now leverages Confluent’s fully managed connectors to enable rapid, low-maintenance integrations across its banking ecosystem, giving the company access to a broader ecosystem of pre-built integrations. These connectors have streamlined processes, allowing the platform team to provision connections quickly, reduce operational overhead, and respond efficiently to business needs, ensuring reliable data flow for critical banking operations. This has accelerated integration timelines by 40-50%, while significantly reducing maintenance burdens on Stone’s engineering teams.

  • Schema Registry: The Schema Registry is a cornerstone of Stone's Confluent ecosystem, especially for its data team. It ensures compatibility, observability, and data integrity across its banking applications by standardizing data formats for real-time financial transactions and internal, event-driven communication. By enforcing governance, the Schema Registry simplifies compliance with banking regulations, reduces debugging time, and enhances data reliability—all critical functions for maintaining a unified, trustworthy data pipeline and enabling the team to focus on delivering value rather than resolving compatibility errors.

Measurable Results: Data that Keeps Business Flowing

Confluent Cloud's architecture directly addresses Stone's most critical technical challenges—reducing operational complexity, ensuring data consistency at scale, and enabling real-time data processing capabilities crucial for business value.

Confluent provides features that streamline Stone’s banking operations, including: 

  • Core Messaging Backbone: Confluent Cloud’s high availability and scalability provide a robust messaging backbone, enabling real-time, fault-tolerant communication that supports its ecosystem’s efficiency and reliability.

  • Real-Time Financial Transactions: Confluent Cloud’s low-latency, fully managed platform ensures seamless, secure transaction processing, which is particularly important for Brazil’s PIX instant payment system. This is vital for customer trust and regulatory compliance.

  • Platform Standardization: Confluent Cloud unified Stone’s messaging infrastructure, simplifying operations and embedding governance for compliance. This reduced complexity, lowered costs, and allowed the company to focus on delivering innovative banking solutions.

All of these technical improvements from Confluent Cloud translate directly into significant business advantages for Stone, including: 

  • Faster Provisioning: Confluent Cloud has streamlined the provisioning of connectors, transforming a process that once took hours into a matter of minutes. This enables rapid deployment of over 100 connectors for its "data bus" initiative, significantly accelerating data integration and boosting operational efficiency.

  • Reduced Cognitive Load on Platform Teams: The fully managed platform automated cluster maintenance, upgrades, and governance from Confluent frees Stone’s teams to focus on business innovation and customer-facing features, while significantly lowering operational complexity.

  • Higher Availability and Resilience: Confluent Cloud’s high availability and scalability ensure robust, low-latency performance, enhancing reliability for critical financial transactions and maintaining customer trust.

“What makes our work particularly valuable is that we don't just maintain pipes for data, we're actively reducing time-to-market for new features and services. By providing a robust, self-service streaming platform, we've empowered product teams to build and deploy faster without getting bogged down in infrastructure concerns. This translates directly to business impact—faster feature releases, improved system reliability, and the ability to scale operations seamlessly during peak transaction periods,” says Oliveira. 

What’s Next?

Stone is actively exploring how real-time data streams from Confluent Cloud can power future AI applications, particularly for advanced fraud detection and personalized customer service. As real-time, event-driven architectures accelerate and AI converges with streaming, Stone is poised to swiftly adopt these trends. 

Confluent provides the clarity and speed needed to scale Stone’s platform and drive future business growth. “We can now deploy services faster, respond to incidents more effectively, and support mission-critical systems with higher confidence. Confluent Cloud has unlocked flexibility and efficiency that wasn’t possible before,” states Oliveira.

Learn more about Stone Pagamentos

Get Started
With Confluent Today

New signups receive $400 to spend during their first 30 days.

See more Customer Stories

Citizens Bank

Citizens Bank

As Citizens Bank modernizes its infrastructure for AI and ML integration, Confluent’s data streaming platform offers an enterprise-grade solution.

Confluent Cloud
Pollinate logo

Pollinate

Data in Motion: Powering Pollinate's Core Business with Confluent

Confluent Cloud
logo-KeyBank

KeyBank

KeyBank Fuels National Digital Bank Initiative with Data in Motion and Confluent

Confluent Platform