Examples

Working examples showing different Orchid capabilities and integration patterns.

Orchid Examples#

This page lists the example projects that demonstrate Orchid's capabilities. Each example is a self-contained project showing different patterns and integrations.

Examples Repository#

All examples are now hosted in a dedicated repository: orchid-examples.

The examples repository contains:

  • Ready-to-run configurations (orchid.yml and agents.yaml)
  • Custom agent implementations
  • Built-in tool definitions
  • Storage backends (in-memory built-in; SQLite via orchid-storage-sqlite; PostgreSQL via orchid-storage-postgres)
  • Identity resolvers
  • Startup hooks and event producers
  • Test suites for each example

Data pipelines#

Two examples are dedicated to the typed data-pipeline feature:

  • Data Pipeline — a direct DAG with deterministic, LLM, agent, decision, map, sub-pipeline and emit jobs. It produces a typed customer-health report and runs from Python, HTTP and the CLI.
  • Event-Driven Data Pipeline — the same pipeline started by a bearer-authenticated webhook and a cron schedule, with SQLite event storage, act_as_user identity and chat-bound results.

Read the Data Pipelines guide for the YAML, shapes, job kinds, result contract and limits.

Running Examples#

Each example can be run in three ways:

  1. Docker Compose — Full stack with API, Qdrant, and dependencies
  2. Standalone API — Run with uvicorn and local Ollama
  3. CLI — Interactive or scripted chat sessions

See the individual example pages below for specific instructions.