Product Tour
See Kuetix and WSL in action
A Paradigm Shift in Engineering
WSL is the engineering language for AI-generated systems.
Traditional Engineering
Human → Code → Runtime
AI code is unpredictable, hard to audit, and unsafe to execute without review.
AI Engineering with WSL
Intent → WSL → Engine
WSL is the safe, validated intermediate language between AI and execution.
"WSL is like SQL for workflows — a standard language between intent and execution."
Why WSL for AI Systems
WSL is a future language for AI systems — not just another DSL.
Declarative Intent Language
Describe what should happen, not how. WSL is not imperative code — it is pure intent.
Deterministic Parsing
AST-friendly grammar with no ambiguity. Every WSL program parses identically, every time.
Validation-First
Schema validation is built into the language. Invalid WSL is rejected before execution.
Separation of Definition & Execution
WSL defines the workflow. The engine executes it. Clear boundary between logic and runtime.
Safer Than AI Writing Code
AI generates WSL, not executable code. No runtime surprises, no hidden side effects.
Debuggable & Explainable
Every transition, state, and action is declared. Easy to audit, trace, and reason about.
How It Works
From intent to execution in five steps
Features
AI-Native Language
WSL was designed from the ground up as the output language for AI systems.
Deterministic Parsing
No ambiguity in the grammar. Same input always produces the same AST.
Schema Validation
Every workflow is validated before execution. Invalid programs are rejected early.
Engine Execution Model
The Kuetix engine provides a sandboxed, observable execution environment.
Integration Actions
Connect to any API, service, or database via declarative action bindings.
Observability
Every state transition is logged and traceable. Full execution history.
Workflow Packages
Share and reuse workflows via the Kuetix package registry.
Portable
WSL workflows run identically across environments and ecosystems.
WSL in Practice
Clean, readable, AI-generatable workflow definitions.
- Declarative state machine syntax
- Typed action bindings
- Built-in delay and retry primitives
- Expressive variable references
- Clear success/error transitions
workflow invoice_reminder {
start: CheckOverdue
state CheckOverdue {
action invoice/query.GetOverdue() as OverdueList
on success -> SendReminder
on error -> Done
}
state SendReminder {
action whatsapp/notify.Send(
to: $OverdueList.contact,
message: "Invoice overdue by {$OverdueList.days_overdue} days"
)
on success -> Wait
}
state Wait {
delay 3d
on success -> CheckOverdue
}
state Done {
action services/common/response.ResponseValue(statusCode: 200)
end ok
}
}
Use Cases
Invoice Automation
Automate invoice reminders, payment tracking, and reconciliation.
SLA Tracking
Monitor service levels and trigger escalations automatically.
Onboarding
Orchestrate multi-step user onboarding with branching logic.
Alert Triage
Route alerts to the right team with AI-generated triage workflows.
Data Pipelines
Build reliable ETL pipelines with declarative state machines.
Compliance
Enforce compliance checks and audit trails via validated workflows.
Roadmap
Spec & Grammar
WSL language specification and formal grammar definition.
CLI
kue CLI for compiling, validating and running WSL workflows.
Package Manager
pkg.kuetix.com — share and reuse workflow packages.
UI Editor
Visual WSL editor with live validation and preview.
IDE Plugins
VS Code, JetBrains, and Neovim plugins for WSL.
Marketplace
Community marketplace for WSL workflow templates.