AI Engineer

PepsiCo taps Traversal to build an agentic mission control. TABLE OF CONTENTS At a glance. Global food and beverage leader PepsiCo operates one of the world's most complex supply chains, with over 3000 applications supporting manufacturing, distribution, and customer operations across multiple geographies. Its OnePepsiCo Operations Centre (OnePOC) team - the first line of defense for production incidents - faced a critical operational challenge: over 80% of Pepsi's major incidents were first detected by employees on the warehouse floor or frontline staff, not by their internal teams. Traversal partnered with PepsiCo to transform their operational command center from a reactive escalation point into a proactive incident prevention engine: a step towards agentic mission control. By automatically triaging and ranking thousands of alerts based on business-context analysis, mapping system interdependencies, and delivering root cause analysis within minutes when incidents occur, Traversal empowered PepsiCo to preemptively identify major incidents and coordinate faster, highly targeted incident resolution across vendor and engineering teams. This resulted in a more than 200% increase in alert processing capacity, freeing up over 500 engineering hours per month across the subset of applications where Traversal was deployed. The challenge. PepsiCo's One PepsiCo Operations Centre (OnePOC) is responsible for safeguarding business continuity across a global, highly interconnected supply chain. Its mandate is not simply to manage alerts, but to detect issues early, coordinate cross-functional response, and minimize operational disruption before it reaches frontline teams. However, operating at PepsiCo's scale introduced structural challenges that compounded across the incident lifecycle: * Enterprise-Scale Signal Complexity: PepsiCo's digital ecosystem generates over 500,000 alerts each month. While only a small fraction correspond to business-impacting incidents, each signal must be evaluated to determine relevance and severity. This volume resulted in sustained alert backlogs - including tens of thousands of unresolved alerts and hundreds flagged as high priority - creating systemic friction in triage and prioritization workflows. The challenge was not organizational efficiency, but the sheer scale and signal-to-noise ratio inherent in operating one of the world's largest supply chains. * Escalation Across a Distributed Operating Model: When alerts matured into Major Incidents (MIMs), resolution required coordination across multiple internal teams, infrastructure partners, and application owners. Given the distributed and interdependent nature of PepsiCo's environment, isolating root causes often involved cross-domain expertise and time-zone coordination. This extended investigation windows and increased business exposure during high-impact events. * Fragmented Infrastructure Intelligence: PepsiCo's observability landscape spans multiple specialized platforms, each providing valuable telemetry but requiring manual correlation to build a complete operational picture. Even understanding the blast radius of an issue required navigating disparate tools and query languages. The limitation was not a lack of data - it was the absence of unified, contextualized infrastructure intelligence aligned to business impact. Its deployment. PepsiCo's pilot phase focused on proving value across 4 core business-critical applications which tied most directly to revenue generating operations. Traversal integrated with PepsiCo's observability landscape, connecting via read-only access to six unique data sources - a mix of legacy and modern platforms: Elastic, AppDynamics, ServiceNow, Azure Data Lake (ADLS), ThousandEyes, and Grafana - all via a SaaS deployment model. The 4-week pilot period was designed with rigorous metrics across three key success criteria: RCA accuracy on both historical and live incidents & alerts, quantifiable MTTR reduction through back-testing, and direct user feedback validation. Over the pilot period, these capabilities demonstrated measurable impact - over 80% RCA accuracy, 500+ monthly engineering hours saved, and elimination of a backlog of 700+ high-severity alerts which helped prevent incidents. Strong user engagement across both OnePOC and MIM teams, validated through direct feedback and careful product analytics, led to approval for strategic expansion from the initial 4 pilot applications to a planned rollout across 66 applications. Traversal's Impact at PepsiCo. Traversal fundamentally transforming how PepsiCo addresses each of its core operational challenges: * Autonomous Alert Triage and Prioritization: In September 2025, Traversal automatically handled thousands of PepsiCo's 500,000 monthly alert volume. With ~10 minutes saved per high-priority alert and at over 100 high-priority alerts each day, Traversal unlocked over 500 engineering hours saved per month, while eliminating the backlog of 700+ open high-severity alerts that required immediate attention. When alerts are now escalated to an engineer, Traversal provides full context - specific infrastructure nodes, root cause hypotheses, and business impact - eliminating reliance on tribal knowledge and enabling faster, informed responses. * Agentic RCA for MIMs: With over 80% RCA accuracy across incidents, Traversal's incident RCA product significantly reduced the time-consuming investigation phase that made incident response costly. Root causes that previously required hours of coordinating across teams are now identified in minutes, reducing downtime and freeing engineering capacity. At PepsiCo, this has translated to an average of 500 engineering hours saved monthly, even with limited scope. * Single Pane of Glass for Infrastructure Intelligence: Traversal provides a single pane of glass across Pepsi's full infrastructure stack, traversing services, dependencies, and telemetry across all observability systems - rapidly retrieving and correlating data to answer questions like "Which components are affected by this incident?" in minutes. This gives OnePOC members and engineers the operational context and response speed typically associated with senior SREs, shifting how PepsiCo's operational teams access critical infrastructure knowledge. Instead of engineers struggling to navigate PepsiCo's complex observability landscape, Traversal enabled natural language queries to interface in a unified manner across all their data sources. Underlying these capabilities is Traversal's Production World Model(TM), which compresses and re-indexes PepsiCo's fragmented observability, dependency, code, tribal knowledge, and infrastructure data into a unified, machine-readable model of the entire production environment. Traversal's Causal Search Engine(TM) then investigates over it, testing thousands of hypotheses in parallel to arrive at a causally consistent diagnosis of likely root cause, affected services, and business impact. Traversal has helped across all of PepsiCo's mission control operations and is deeply embedded in daily workflows. The business impact is substantial: investigation time has been reduced by 70%, enabling faster resolution and materially reducing operational friction during incidents, freeing up engineering capacity. Inside a real incident. At 4:06AM on October 31, 2025, warehouse workers in one of PepsiCo's main distribution centers began experiencing failures with a critical application: the technology that tells workers which pallets to move and where. When warehouses can't move inventory, distribution capacity drops, and manufacturing plants lose storage space for finished goods. The AppDynamics alert triggered immediately. Traversal automatically launched an investigation. Within 3 minutes, it identified the root cause: CPU saturation on a shared middleware node, correlated with a critical application deployment 21 minutes earlier. Traversal mapped the impact to co-hosted services, assessed business criticality, and recommended response escalation to the Platform Operations team. OnePOC escalated with full context, and PepsiCo's infrastructure partner knew exactly which host to inspect and what to fix. The incident was contained within 30 minutes-something that would've taken hours for the initial manual investigation alone. Towards an agentic mission control. PepsiCo's deployment of Traversal represents more than operational efficiency gains: it's a fundamental architectural and conceptual shift towards an agentic mission control. Building this future requires solving all three of PepsiCo's core operational challenges - and Traversal addresses each one. Alert triage, incident RCA, and seamlessly understanding your infrastructure, which once required extensive manual effort now occur autonomously with minimal human intervention. Hence, enabling a small team of human orchestrators to efficiently coordinate complex multi-team responses, and empowering senior engineers to focus on system design and long-term resilience rather than reactive firefighting. As the system learns from resolution patterns, the goal is to move towards self-healing where common issues are automatically remediated before human intervention is needed. To see Traversal in action, book a demo today. "Operating at PepsiCo's scale - 3,000+ applications, hundreds of thousands of alerts - requires intelligent automation beyond traditional monitoring. Traversal's AI SRE agents cut through this enormous complexity, automatically triaging alerts and surfacing root causes in minutes rather than hours. What previously required coordinating 20+ engineers across time zones now happens autonomously. It's transforming our operations from reactive firefighting to preventing incidents before they reach the warehouse floor - the future of reliability at global scale." Vinod Chilakalapudi - Director of IT Operations, PepsiCo

Traversal, an AI lab building agents for enterprise site reliability engineering, has announced six senior leadership hires across go-to-market and engineering within a single month. The company's headcount has grown to over 90, representing a 110 per cent increase in six months. New appointments include Jim Cavanaugh as SVP of Worldwide Sales, Ryan Powers as SVP of Marketing, Patrick Wade as VP of Worldwide Field Engineering, and Maxime Petazzoni as Head of Engineering. The hires bring experience from companies including Cribl, Redis, SignalFx and Splunk. The expansion follows Traversal's recent investment from Amex Ventures and deployment across American Express. A Fortune 100 financial services case study showed 32 per cent reduction in potential mean time to resolution and 82 per cent root cause analysis accuracy.