Reinventing SRE with AI: Our Investment in Cleric

We’re excited to share that Vertex Ventures US has led a new $5.5M seed round in Cleric, bringing total funding to $9.8M alongside our friends at Zetta Venture Partners. You can find Cleric’s own announcement here, and go deeper on the news with SiliconAngle’s story here.

Cleric is building an AI SRE that turns every investigation into operational memory — investigating like a senior engineer, giving teams SRE-level support, and reducing time to resolution and operational impact.

The net result: SREs spend less time constantly putting out fires and more time delivering real business value. In practice, mature teams like BlaBlaCar are already using Cleric to move from chasing one alert at a time to building intelligent coverage across services. Engineering leaders tell us Cleric surfaces the patterns behind recurring failures, which lets them eliminate entire classes of issues. That shift gives engineers meaningful time back, reduces repeat root causes, and, in early deployments, has reclaimed roughly 20–30% of total engineering capacity.

We met Shahram Anver and Willem Pienaar, the founders of Cleric, early in their journey. They built and operated large-scale infrastructure at companies including Gojek, giving them an invaluable first-hand view into how AI-driven disruption necessitates change in the way that we monitor and manage the entire IT stack.

AI for production is fast becoming its own category. Over the last year, my colleagues at Vertex and I have dug into how AI is changing software not only in how it’s written but in how it runs. Across the IT stack, we’re moving from manual, human-paced workflows to systems that understand context and surface the next right action. Provisioning, CI/CD, security, observability and performance, and incident response are all being rethought for environments that are too complex and too dynamic to manage with dashboards and scripts alone.

That shift is happening because modern applications are no longer predictable. AI-generated code, agentic workloads and non-deterministic behavior introduce cascading changes, far more telemetry, and a whole new level of complexity. A single user action can trigger dozens of downstream processes, scale infrastructure in seconds and mutate environments in real time. Teams using traditional tools will struggle to keep up.

Observability is where this pressure shows up first. Most issues never become headline incidents, yet they still consume hours. An engineer is deep in a refactor, an alert fires, a Slack thread starts, logs and metrics are pulled, a deployment diff is checked, the immediate problem is fixed, and the context disappears. A few weeks later, it happens again. The investigation repeats. Anything learned rarely gets preserved, especially in the inevitable case of SRE staff turnover. Ultimately, time and velocity are lost.

This is where Cleric is different. It mirrors how experienced engineers actually investigate production issues. It asks what changed, correlates logs, metrics and deploys, explores multiple hypotheses in parallel and presents evidence-backed findings in Slack where engineers already work. Every new investigation makes Cleric smarter, as day-to-day triage becomes operational memory that compounds. The next time a similar issue appears, the organization doesn’t start from zero: Cleric already knows where to look.

Ultimately, we believe that complexity will keep rising faster than human attention can keep up. Systems like Cleric that capture and reuse investigative knowledge will become foundational to how modern software is run.

We are proud to partner with Shahram, Will and the team as they scale from strong product usage to broader adoption.