Powering index generation for global markets
STOOX produces and maintains a global family of more than 18,000 indices, used by institutional investors to benchmark portfolios and power financial products such as ETFs, derivatives, and structured investments. These indices are embedded in decision-making processes across pension funds, asset managers, and trading desks, where accuracy, consistency, and availability are critical requirements.
At the core of this responsibility is the platform that generates index data. Built on Google Cloud, it now supports 100% of STOXX’s index calculation and data processing workloads, combining large-scale data pipelines with compute-intensive calculations across services such as BigQuery, Cloud Run, and Cloud SQL.
As the platform evolved, so did the expectations placed on it. STOXX needed to ensure that its systems could deliver consistent performance while providing the transparency required in a highly regulated environment. “As we scaled our platform in the cloud, we needed a way to fully understand how our systems were behaving in real time,” says Mit Saru, Senior Director of Platform Engineering at STOXX. “When you are responsible for data that drives investment decisions, you cannot rely on assumptions. You need clear, provable insight into performance and reliability.”
Scaling visibility alongside cloud transformation
The transition to Google Cloud marked a significant step in STOXX’s modernization strategy. The team migrated legacy systems, built new cloud-native services, and established a structured environment aligned with Google best practices. This transformation enabled greater scalability and flexibility, but it also introduced new levels of complexity across distributed services.
“When you are responsible for data that drives investment decisions, you cannot rely on assumptions. You need clear, provable insight into performance and reliability.”
As workloads expanded, the ability to observe and understand system behavior across environments became increasingly important. STOXX needed to correlate signals across applications, infrastructure, and data pipelines to maintain confidence in performance and availability. “We are operating in a regulated space where we must be able to explain exactly what happened in any situation,” says Saru. “That requires a level of visibility and traceability that matches the sophistication of the platform itself.” At the same time, the team needed to manage cloud usage effectively, ensuring that growth in compute and data processing was aligned with cost efficiency.
Establishing a unified observability model
To meet these requirements, STOXX implemented a unified observability model across its platform, bringing together Datadog Application Performance Monitoring, Infrastructure Monitoring, and Log Management into a single, consistent view. This allows teams to correlate metrics, traces, and logs in real time and move quickly from detection to diagnosis.
This model has become central to how STOXX operates. Alerts are generated based on system behavior, and investigation begins immediately within a shared environment. “Datadog is the first place we go when something happens,” says Saru. “It gives us the context we need to understand the issue quickly and respond with confidence.”
Synthetic Monitoring adds continuous validation of critical workflows, ensuring that applications are accessible and performing as expected. This provides an additional layer of assurance that complements internal telemetry.
Cloud Cost Management plays a key role in maintaining operational efficiency. STOXX uses Datadog dashboards during monthly FinOps reviews to track resource consumption and understand cost drivers across its cloud environment. This enables the team to balance performance and scalability with financial control. “We need to understand not only how systems perform, but also how they scale and what that means from a cost perspective,” Saru explains.
Dashboards provide a shared operational layer used by both SRE and Level 1 teams. This shared visibility supports a shift-left approach, enabling earlier detection and resolution of issues while maintaining alignment across teams.
Looking ahead, STOXX is exploring how to extend this model with AI-driven capabilities. The team plans to evaluate Bits AI to support incident investigation by guiding teams toward likely root causes and next steps, further improving speed and consistency in operations. “Bits AI has the potential to take our incident response to the next level by helping teams move faster with greater confidence,” says Saru.
Improving operational confidence and scalability
With a unified observability model in place, STOXX has strengthened its ability to operate at scale with confidence. Teams now have a real-time, end-to-end view of system performance, allowing them to validate that applications are meeting expectations and that data is delivered reliably.
“Bits AI has the potential to take our incident response to the next level by helping teams move faster with greater confidence.”
This visibility has improved how incidents are managed by providing immediate context and reducing the effort required to investigate and resolve issues. It has also enhanced traceability, making it easier to explain system behavior and support regulatory reporting when needed.
Operational efficiency has improved across the platform as well. Engineers spend less time gathering and correlating data and more time focusing on platform improvements and innovation. Level 1 teams are able to resolve a greater share of issues independently, improving responsiveness and reducing escalation.
At the same time, better visibility into cloud usage has strengthened cost awareness and control. With centralized insight into resource consumption, STOXX can make more informed decisions about scaling its platform while maintaining efficiency.
As STOXX continues to expand into new regions and support additional workloads, observability is now embedded as a core part of the platform. “Datadog gives us proof that our systems are performing as expected,” says Saru. “That level of confidence is essential when your data is used to drive decisions across global markets.”
“Datadog gives us proof that our systems are performing as expected. That level of confidence is essential when your data is used to drive decisions across global markets.”
