Project One.
The Concept
Delivering an AI-Native ODA canvas.
Scope & key activities
The scope of the project is to deliver a live demonstration of a running AN Agent on the AI-Native ODA Canvas, showing the end-to-end path from data ingestion and reasoning through operator execution to service recovery. This will serve as the reference implementation for CSPs accelerating AI-native adoption.
The project Members will showcase a live demo at DTW Ignite in June 2026 in Copenhagen; with a Beta version showcased at MWC in February 2026 in Barcelona.
Objectives & outcomes
The objectives are to enable a unique “one-to-many” innovation model and to demonstrate a working reference implementation of the AI-Native Blueprint v1.0 principles.
The expected outcomes include:
- A Kubernetes-based Canvas running on AWS.
- Installed Databricks (with curated dataset) and integrated CKAN catalogue.
- Installed MuleSoft/AWS MCP for A2A interaction.
- Installed AWS Bedrock for model hosting.
- Exposed Canvas operators (Data, Interaction, Model) as microservices.
- Validated operator flow through unit testing with stubs/logs.
- Running Autonomous Network Agent (fault management).
Join this project
Participants




