Co-founder & CEO
Profile summary: PhD in Electrical Engineering and Computer Science at Massachusetts Institute of Technology
Traversal
UnknownTraversal provides an AI-powered site reliability agent that autonomously troubleshoots, remediates, and prevents production incidents in complex enterprise environments. The platform builds a continuously updated causal "Production World Model" of the entire system, investigating thousands of hypotheses in parallel to isolate true root causes. Deployed at companies like DigitalOcean, PepsiCo, and Cloudways, Traversal reports over 90% root cause accuracy and reductions in mean time to resolution of up to 70%.
- AI Agents
- Artificial Intelligence
- Developer Tools
- Enterprise Software
- Software Only
- Data & Analytics
Updated 2 days ago
Funding
Raised to date
$173MRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
across 4 rounds
AV+2$120M · Series B
Announced Nov 2026
Disclosed cumulative: $168M
4/4
RoundDateAmountValuationInvestors
Founders
Co-founder
Profile summary: PhD in Electrical Engineering and Computer Sciences at University of California, Berkeley
Co-founder
Profile summary: PhD in Electrical Engineering and Computer Science at Massachusetts Institute of Technology
Product
Problem
Modern software systems generate vast amounts of telemetry data, including logs, metrics, and code, making it difficult for developers to quickly identify the root causes of errors and latency. Traditional observability workflows are complex, often overwhelming engineers with a flood of alerts and logs, leading to alert fatigue and prolonged debugging times. Downtime costs enterprises up to $1 million per hour during major incidents, and engineers can spend up to 30% of their on-call time firefighting instead of building.
Solution
Traversal provides an AI-powered site reliability agent that analyzes telemetry data to automatically troubleshoot, resolve, and even prevent production incidents. The platform builds a continuously updated, causal "Production World Model" of the entire production system, then investigates thousands of hypotheses in parallel to isolate true root causes—not just symptoms. Traversal parses logs, metrics, traces, and codebases, translating complex data into natural language and surfacing the blast radius, key bottleneck services, and candidate root causes with supporting evidence. The platform autonomously triages alerts, provides natural-language answers about live production state, and can execute automated remediation actions. Traversal integrates with existing observability stacks through read-only access and flexible deployment models, including optional on-prem deployment for data privacy.
Target Audience
Traversal is designed for Site Reliability Engineers (SREs), DevOps teams, and software engineers in enterprises with complex infrastructure and microservices architectures, including large financial institutions, cloud providers, and managed hosting platforms.
Features
- Production World Model: continuously updated, AI-readable causal representation of the entire production system that distinguishes multi-hop root causes from coincidental correlations
- Incident RCA: traces incidents across services, dependencies, and changes to isolate true root cause and remediation path in minutes, with >90% accuracy reported on hundreds of high-impact incidents
- Alert Intelligence: autonomously triages alerts using historical behavior and cross-system relationships, prioritizing by business impact and grouping duplicate fires
- Self-healing: converts diagnosis into automated remediation, compressing recovery time; at Cloudways, achieved 95%+ accuracy on common web hosting issues like DDoS and disk errors
- Code Resilience: feeds production context back into development, flagging risky changes before merge and suggesting fixes based on historical incident patterns
- Production Support: natural-language interface across all data sources, giving engineers answers grounded in real-time telemetry, code, and tribal knowledge
- Read-only deployment by default, without agents, schemas, or sidecars, on top of existing observability tools including Datadog, Prometheus, and Grafana
- Flexible deployment options including on-premise for strict data residency requirements