Gremlin

Gremlin Inc. develops a software platform that lets engineering organizations deliberately inject faults into their systems—latency spikes, storage errors, database congestion, host and network failures—in a controlled way, in order to find weaknesses before they cause customer-facing outages. The company positions this as "chaos engineering" and, more recently, as broader "enterprise reliability management": combining passive risk detection, service and dependency discovery, pre-built and customizable reliability test suites, GameDay coordination, and aggregate reliability scores that are meant to serve as forward-looking indicators in place of lagging metrics such as MTTR, uptime, and incident counts.

The platform is cloud-native and is marketed as supporting AWS, Azure, and GCP, along with Linux, Windows, container environments including Kubernetes, on-premises and bare-metal deployments, service meshes (Istio via Envoy, Linkerd), and serverless platforms such as AWS Lambda through its Failure Flags feature. A Private Edition can be run fully isolated inside a customer network. Safety mechanisms include blast-radius management, halt conditions, role-based access control, multi-factor authentication, single sign-on, and audit trails; the company states it is SOC 2 compliant (SOC 2 Type II certification completed in 2019) and undergoes third-party security audits.

More recent product direction centers on reporting and AI. Reliability Intelligence (2025) adds tailored experiment analysis and remediation recommendations; Disaster Recovery Testing (2026) targets zone, region, and datacenter evacuation and failover exercises; and Foresight AI is being previewed as a capability that scans, tests, fixes, and verifies resilience automatically.

Founding story

The company's timeline traces its origins to Kolton Andrus building fault injection at Amazon in 2009; Andrus later designed Netflix's failure injection service. Gremlin Inc. was founded in January 2016 by Andrus, who serves as CEO, together with CTO Matthew Fornaciari, whom he had worked with at Amazon. Both co-founders served as 'Call Leaders' at Amazon and Netflix, responsible for resolving global outages.

Business model

B2B software sold to enterprise engineering organizations, with a self-serve free trial (described variously as 30-day and 14-day, no credit card required) as an entry point and demo-led enterprise sales. The company also describes an "expert partnership model" in which its engineers embed with customer teams to build out their reliability practice.

Sources describe a commercial SaaS platform with free trials and enterprise deployments, including a self-hosted Private Edition, but do not disclose pricing or revenue figures.

Traction

Publicly cited traction consists of enterprise customer claims and outcome statistics rather than financial disclosures: use by four of the five largest US banks, a 50% downtime reduction at a major US insurer, a 90% reduction in disaster-recovery testing time at a top-5 global bank, 60 critical failure modes identified at a top-5 US bank serving 100 million customers, and 99.99% availability achieved on a platform migration. The certification program issued more than 2,000 credentials in its first year.

Latest developments

The website prominently previews Foresight AI, an early-access capability that scans and tests systems for potential failures, applies fixes, and verifies resilience. The company timeline lists Disaster Recovery Testing (January 2026) for zone, region, and datacenter evacuation testing, following the November 2025 Dynatrace partnership and the August 2025 launch of Reliability Intelligence. Recent third-party mentions listed by the company include Tech Times, The Prime View, and SD Times items dated mid-2026.