Agency Orchestrator
Self-hosted AI agent orchestrator builds teams and visual workflows, with Ollama for local inference or cloud APIs for hosted models.
Alternative to:
Agency Orchestrator coordinates specialist AI roles for developers and solo builders working on product plans, research or content. A task description can become a team that divides the work and combines its results.
- Team selection: Let AI choose specialists, select roles yourself, or use workflow templates for product planning, content production and investment research.
- Visual workflows: Edit tasks, roles and their connections on a canvas, then follow progress and intervene through the local browser interface.
- Deliverable checks: Workflows can request human sign-off and verify results against acceptance criteria, while independent tasks run in parallel and later steps receive earlier outputs.
- Revisions and reports: Resume runs, provide feedback for revisions and export reports to share the work.
The software is open source under Apache 2.0, with local browser, desktop and Docker deployment options. Ollama supports local model inference. DeepSeek and OpenAI-compatible APIs send model requests to the selected provider; authenticated coding tools such as Claude Code and Codex can also act as providers using an existing login rather than a separate API key.
Claude Code and Cursor can invoke workflows through MCP. Generated artifacts are saved in local output and project directories.
Logotyper
