Kumeo Documentation

Note: Kumeo is currently in early development

Welcome to the Kumeo documentation. This project is actively being developed, and we’re excited to share our vision for orchestrating distributed agent workflows.

What is Kumeo?

Kumeo (from kume, meaning “together” in Mapudungun) is a domain-specific language (DSL) designed to define complex workflows where heterogeneous agents collaborate via events. It employs a multi-language compilation approach, generating optimized code for each agent type (Rust for performance-critical components, Python for ML and Bayesian operations), uses NATS for event streaming, and deploys to Kubernetes for scalability.

Project Overview

Kumeo aims to solve the challenge of connecting different AI and ML components into cohesive, event-driven systems. Key aspects include:

Key Features

  • Declarative Workflows: Define agent interactions as event-driven flows using a simple, readable syntax
  • True NoOps Solution: The language itself handles everything from code generation to deployment - just write your workflow and run it
  • Multi-Language Runtime:
    • Python: For rapid development, ML, and data science workloads
    • Rust: For high-performance, low-latency components
    • Seamless Interop: Agents in different languages can communicate transparently
  • Agent Types: Support for LLMs (Ollama/OpenAI), ML models (scikit-learn, ONNX), Bayesian networks, and human-in-the-loop
  • Event Orchestration: Built on NATS for real-time, distributed communication
  • Kubernetes Native: Auto-generates deployment manifests for scalable infrastructure
  • Visual Editor: Svelte-based UI for workflow design and monitoring

Project Structure 

kumeo/
├── compiler/       → Rust-based compiler (Kumeo → Rust + Kubernetes YAML)
├── runtime/        → Agent execution engine (NATS integration)
├── ui/             → Svelte visual workflow editor
├── examples/       → Sample Kumeo workflows
└── kubernetes/     → Deployment templates

Development Status

This project is in active development, with no executable releases available yet. We’re following a structured implementation roadmap to build the core components:

  1. Foundation Phase: DSL design, compiler foundations, runtime basics
  2. Core Implementation: Agent types, code generation, Kubernetes integration
  3. Advanced Features: Subworkflows, distributed tracing, state persistence
  4. UI & User Experience: Visual editor, monitoring dashboard, documentation
  5. Testing & Refinement: Test framework, example workflows, beta program
  6. Production Release: Final documentation, v1.0 release

Documentation Sections

As the project develops, this documentation will expand to include:

Thank you for your interest in Kumeo as we work to create a powerful platform for agent orchestration!