Architecture
The Composer AI Project includes both the code base and a working tech stack for setting up the base generative AI solution, including:
- Composer AI Frontend Application (PatternFly)
- Conductor (Quarkus API Server)
- LLM Serving Runtime (Granite installed on vLLM)
- Vector Database (Elastic)
- Example Ingestion Pipeline (Tekton/KFP)
This allows for developers to be able to quickly deploy the application. Note if you have access to the Red Hat demo platform, there is a pre-provisioned demo that creates a fully functional instance of the application on demand.