American Express Taps Traversal to Transform Site Reliability Engineering with AI

Traversal, the frontier lab building AI agents for enterprise-grade site reliability engineering (SRE), today announced a strategic partnership with American Express that includes an investment from Amex Ventures and the deployment of Traversal’s agentic SRE platform within American Express’s global technology infrastructure.

“We’re excited to deepen our partnership with American Express as we tackle one of the toughest problems in software: helping engineers troubleshoot complex incidents across distributed infrastructure at enterprise scale,” said Anish Agarwal, co-founder and CEO of Traversal. “American Express is setting the standard for operational resilience in an era that demands it.”​​​​​​​​​​​​​​​​

For most companies, a website outage costs an estimated $1.9 million per hour and takes 77 hours to remediate. For American Express—a multibillion-dollar global payments leader processing billions of transactions daily—the calculus changes by an order of magnitude. Knowing AI-generated code is making enterprise systems increasingly complex, more prone to incidents, and harder to troubleshoot, American Express is proactively investing in a fundamentally new approach to operational resilience.

Within six months of deployment, Traversal delivered an 82% root cause analysis accuracy rate, a 32% reduction in potential mean time to resolution (MTTR), and the ability to autonomously trace root cause in minutes for incidents that previously required extensive cross-team coordination. The platform ingests 250 billion logs of interest daily across American Express's infrastructure, which spans Kubernetes clusters, Lambda functions, mainframe systems, and traditional on-premises environments. 

Traversal leverages its founding team’s years of PhD research in agentic AI and causal machine learning at MIT, Columbia, and Cornell to autonomously detect, troubleshoot, and remediate complex production incidents. At its core is the Production World Model™, a continuously and autonomously learning, machine-readable model of an enterprise’s production environment. It compresses and re-indexes raw telemetry and code into a unified structure purpose-built for AI reasoning at scale. On top of that model runs the Causal Search Engine™, which executes directed investigations. It eliminates hypotheses inconsistent with system topology and behavior, converging on a single, evidence-backed root cause. 

Traversal is built with a security-first, read-only architecture and a flexible deployment model — including full on-premise support — giving customers complete control over sensitive data. It is trusted by leading enterprises to operate inside mission-critical environments.

Based in New York City, Traversal emerged from stealth in June 2025 with $48M in funding led by Kleiner Perkins and Sequoia.

“Both consumers and enterprises have increasingly high expectations for performance and reliability when it comes to the technology tools they use. That’s why AI-driven site reliability engineering is becoming mission-critical,” said Kevin Weber, Managing Director at Amex Ventures. “Traversal’s security-first architecture, enterprise-ready deployment model, and work in causal ML and AI agents position them to emerge as a leader in AI-powered SRE.”

To learn how Traversal supports enterprise-scale reliability, book a demo today.