Case Study: Captur achieves proactive crash detection and cost-effective SDK scaling with Datadog RUM & Workflow Automation

Captur scales SDK quality and customer trust with Datadog RUM and Workflow Automation

Captur is a London-based AI platform that performs real-time, on-device image verification via an SDK embedded in customer apps. Because the SDK runs inside diverse host apps and environments, crashes from the host were often misattributed to Captur, namespace collisions and unsampled session floods made root-cause analysis slow and costly, and investigations could take a day or more—creating customer frustration and blocking scale.

By adopting Datadog Real User Monitoring and Workflow Automation, Captur gained encapsulated SDK visibility, dynamic sampling without redeploys, and automated crash triage. The team now filters and samples sessions to capture 100% of crashes while keeping normal-session sampling low, excluded 100k unrelated crash sessions from one client, reduced irrelevant session ingestion by over 75% (4M→<1M), and scaled to hundreds of thousands of daily sessions with only ~2x cost—cutting troubleshooting time from hours to seconds and enabling proactive customer outreach.

Captur

Sumanas Sarma

Chief Technology Officer