Prequel

Prequel offers a distributed detection platform for infrastructure and security teams. Rather than shipping telemetry to a central store for analysis, the platform streams detection rules to where data already resides and evaluates them in place, which the company positions as a way to avoid data pipelines, cloud egress charges and storage costs. Detections cover application code, message queues, databases, container runtimes and Kubernetes objects, and are built from what the company calls global failure knowledge — bug reports, community discussions and post mortems — codified into an extensible library of problem detectors spanning open source bugs, misconfigurations and software antipatterns.

The commercial product is built on top of the open source preq project, and extends it with distributed detection, broader Common Reliability Enumeration (CRE) coverage, workflows, exclusive intelligence and production-grade integrations. Detections carry a CRE identifier, severity, impact and mitigation scoring, and suggested remediation, and can trigger automated runbooks such as Slack notifications, Jira issue creation, executables and CronJobs. Recent releases added detection grouping and filtering across services, a dismissal workflow, improved log investigation views, and new data sources including journald logs, kubelet metrics and logs, and Kubernetes configuration and workload data. Deployment has been consolidated into a single public container image distributed via DockerHub.

The engine supports declarative rule syntax, event sequencing and correlation, and negative conditions, allowing rules such as detecting a specific ordered sequence of log events within a time window unless a given event also occurs. The company describes the distributed engine as patent-pending.

Business model

Software sold to enterprises on a flat-rate subscription basis with no volume-based data fees; a free tier and a 30-day trial are offered, with self-serve sign-up alongside sales-led demos. An open source detector, preq, serves as the entry point to the commercial platform.

Flat, predictable pricing based on node count rather than data volume, with three published tiers — Startup (up to 100 nodes), Scale Up (100-500 nodes) and Enterprise (500+ nodes) — all priced on request.

Traction

The open source preq repository has 343 stars, 28 forks and 126 commits. The commercial platform has iterated through releases up to v0.14, and hundreds of new CREs have been added by the community and the internal research team. The company states it has assembled what it describes as the largest reliability intelligence library and cites customer-referenced outcomes including 73% faster detection, $1.2M in log cost savings, 26x more coverage, 13 engineer-hours saved per week, 67% less impact and 82% less manual analysis; these figures are company-reported.

Latest developments

Release v0.14 introduced cross-service detection grouping, a detection dismissal workflow, new data sources (journald logs, kubelet metrics and logs, Kubernetes configuration and workload data), a consolidated public container image and migration of official images to DockerHub. Earlier release v0.12 shipped Engine v2.0 with sequencing, correlation and negative conditions, impact and mitigation scoring, community CRE access with daily rule updates, detection grouping and filtering, and improved log investigation. Recent blog posts address the platform's detection engine architecture, 2026 plans for its reliability intelligence library, and the March 2026 end-of-maintenance for Ingress NGINX.