wtf is auto-workflow-orchestrator?

AutoWorkflow Orchestrator is an open-source platform for automating complex business workflows inside large companies, aimed at industries like automotive, manufacturing, and logistics. The project positions itself as a cheaper alternative to expensive commercial tools such as SAP Build Process Automation or Siemens Teamcenter. The system is built around five specialized AI agents that each handle a different job: taking in and validating new orders from channels like email or a customer portal, managing change requests across multiple business systems, tracking product lifecycle changes such as engineering change requests, managing the lifecycle of SIM cards for connected vehicles, and monitoring different company systems to catch when their data drifts out of sync with each other. A supervisor component coordinates these five agents, and a separate compliance layer handles GDPR requirements like consent tracking and data subject requests. Under the hood, the backend is written in Python using FastAPI, with the agents built on LangGraph and connected to a large language model through OpenRouter. Data is stored in PostgreSQL, background jobs run through Redis and Celery, and the frontend is a React and TypeScript application. The whole stack can be run locally with Docker Compose, which also sets up monitoring dashboards through Prometheus and Grafana. To try it, you need Docker, Python 3.11 or newer, and Node.js 18 or newer. After cloning the repository, you start the database and Redis with Docker Compose, set up a Python virtual environment, run database migrations, and start the backend and frontend separately, or run everything at once with a single Docker Compose command. Once running, interactive API documentation is available through a built-in Swagger interface. The project is released under the MIT license and is built by a group calling itself AgentsFactory, described as an autonomous AI swarm.