inisiyatif / labs / zervanor

Production Ready · v2.0.24

Connect agents, apps,
and LLMs
with one event standard.

Route events between apps and agents. Deploy in seconds. Observe everything.
No vendor lock-in. CloudEvents-first. Kubernetes-native.

Visit zervanor.com Discuss with Inisiyatif
WhatsApp → LLM → HubSpot | Webhook → Router → Slack | Email
zervanor-cli — production shell
➜ ~ zervanor init my-agent --template webhook
✔ Initialized project my-agent
✔ Created manifest.json from template
➜ ~ zervanor push --deploy
ℹ Validating manifest... OK
ℹ Building container... DONE (1.2s)
ℹ Deploying to K8s... DONE (12s)
✔ Agent is running. Live trace is online.
PLATFORM PREVIEW

Code for developers.
Visual for everyone else.

CLI and SDK for developers. Visual workflow builder and admin dashboard for operations teams.

Visual Workflow Builder
Trigger
Process
Action
Admin Dashboard
12.4kEvents / hr
99.97%Uptime
48Active Agents
AgentStatusEvents
order-processorRunning3,241
support-botRunning1,892
invoice-syncScaling567
CORE FEATURES

Built for shipping,
not for demoware.

CloudEvents First

Every source, from WhatsApp to Stripe, is normalized into one vendor-neutral event protocol.

Native CE 1.0

AgentSDK & CLI

Build agents in TypeScript, push with one command, and deploy into isolated, Kubernetes-native workloads.

init → push

Visual Workflow Builder

Design multi-step workflows with trigger matching, conditional routing, and auto-layout for operations teams.

15+ node types

Marketplace & Ecosystem

Install ready connectors and agent templates in one click, then publish your own as reusable runtime modules.

500+ integrations target

K8s-Native Security

Every agent runs in its own pod with namespace-level multi-tenant separation, network policies, and secret management.

Pod-level isolation

Full Observability

Trace every workflow step with structured logs, live metrics, and distributed telemetry for failure analysis in seconds.

OpenTelemetry-ready
RUNTIME MODEL

Two runtimes.
One platform.

Lightweight HTTP connectors for APIs. Isolated Kubernetes pods for long-running AI agents. Orchestrated together through CloudEvents.

Connectors

  • Stateless HTTP execution
  • Low-latency API integrations
  • REST, GraphQL, webhooks

Stripe · HubSpot · Gmail · Jira

Event Router CloudEvents backbone

Agents

  • Isolated Kubernetes pods
  • Long-running, stateful workloads
  • LLM, messaging, and orchestration logic

Claude · Gemini · WhatsApp Bot · Custom AI

ENTERPRISE GOVERNANCE

Governance your security team
will actually approve.

Multi-Tenant Isolation

Each organization runs in its own Kubernetes namespace with separate workloads, secrets, and data boundaries.

RBAC & SSO

Fine-grained access policies integrate with existing identity providers via SAML or OIDC.

Audit Logs

Every deployment, config change, and privileged action is logged and queryable for compliance review.

WHY NOW

Why now?

The automation market is fragmented. Zervanor unifies the stack.

01

AI agents need an operating layer

LLMs can produce actions, but enterprises still need deployment, routing, observability, and governance.

02

Stop paying the per-step tax

A 10-step workflow should not behave like 10 unrelated billable tasks. Zervanor treats a workflow as one execution unit.

EARLY ACCESS

Be first to build on Zervanor.

Priority onboarding, template starter kits, and a direct path from prototype to production-grade automation.

Open zervanor.com Contact Inisiyatif