Sansan, Inc. Detects Issues Faster, Resolves Production Issues Within 30 Minutes

Maximizing observability for better uptime

Bill One needed a solution that would make system-wide observability possible. “We want to visualize the entire system through a service map and see what happens at any specific time by drilling down to the data to improve our business agility,” says Maeda. “We also expect a high data sampling rate for conducting detailed analysis.” To avoid vendor lock-in, the new solution should also comply with OpenTelemetry, an open-source standard for collecting telemetry data for system visualization.

As it turned out, Splunk’s solutions suited all these needs. Bill One’s team used Splunk Application Performance Management (APM) to create a monitoring environment that would fit its observability goals, and Splunk Infrastructure Monitoring (IM) to gain comprehensive infrastructure performance visualization to boost business efficiency and agility — for example, the full-stack visibility allows Bill One’s team to detect all system failures and incidents, which was unachievable before using Splunk. Full compliance with OpenTelemetry also enabled the team to skip the technical conversion process required by alternative solutions. “Splunk’s 100% sampling rate enables us to oversee the health of Bill One and accommodate our business growth at rapid speed,” Maeda adds. Splunk’s account teams also closely partnered with the teams at Sansan to make sure the solutions fit its requirements. “Splunk has a solid understanding of our business,” says Maeda.“They are even kind enough to walk through the Splunk dashboard with us step-by-step.”

With built-in visualization and AI-driven workflows, Splunk APM has given Sansan’s teams an extensive view across the Bill One system and enables cross-team operations to investigate issues quickly. “It’s particularly useful during peak periods when there are lots of invoice exchanges,” Maeda explains.

Bill One processes hundreds of exabytes of traffic data monthly on Cloud Run, a serverless container runtime environment. Splunk APM has helped the team reduce latency and maximize uptime of the service, in giving them a means to visualize microservices with full-fidelity tracing and streaming analytics. “We can also use the detectors to dynamically monitor alerts and set latency thresholds for endpoints that would impact the service,” says Maeda. “It’s great to see that more colleagues are able to check latencies themselves and make improvements to boost the quality of service for the end users.”