Scaling AI Development With Software Templates in Backstage: Challenges, Lessons, and Integrations
summary
Shows a Backstage software-template system that scaffolds AI apps with Tekton pipelines, ArgoCD GitOps and a Model Catalog, and demonstrates three concrete integration modes: starter models (bootstrapped Llama.cpp + Granite model images), bring your own model (external model server URL with auth), and catalog model (Backstage catalog entries bridged from a model registry). Templates auto-create repos, run pipelines (all green in demo), patch deployments, and launch a Streamlit app; Renovate is used to automate image updates.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:03Why create AI templates
- 3:36Template architecture
- 6:30Model integration strategies
- 11:48Live demo: three strategies
- 17:15Challenges and UX issues
- 19:01Resources and demo reproduction
- 20:25Summary and Q&A
related talks
48mThe Room of Requirements for Developers: Scaling AI Magic with Internal Portals
01h 29mRunning AI Application in Minutes w/ AI Templates
21mLeveraging GenAI for Seamless Developer Experiences in Backstage
55mBuilding and scaling AI applications for Data Scientists and Developers
13mAI App Platform: Elevating the Dev Experience
06m⚡ Lightning Talk: Supercharge Agentic AI Apps: A DevEx-Driven Approach To Cloud Native Scaffolding
01h 03mAI Templates
26mWelcome to the AI Jungle! Now What?