Fine-Grained Authorization & Beyond: Mastering Keycloak Patterns in AI-infused Apps
summary
Claims that Keycloak Authorization Services can enforce resource-and-scope permissions for AI model access and that Quarkus supplies a built-in filter for token propagation between services. Demo consists of two Quarkus apps acting as MCP client and MCP server, using GitHub OIDC for login, Postgres for mapping custom permissions, and Google Gemini as the model; the poem-generator prototype mostly works but shows a live-demo hiccup. Notes multi-tenancy options, custom Keycloak policies, and Java-focused toolkit LangChain4j to enable fine-grained authorization across many models (cites ~1.7M models).
AI-written summary · about
chapters (6)
- 0:00Introduction and challenges
- 2:03Security architecture goals
- 4:54Core security patterns
- 8:06Access control strategies
- 12:46Live demo and implementation
- 23:16Summary and takeaways
related talks
31mSecure your Quarkus Applications
33mSecuring Your Quarkus Applications
50mFine - Grained Authorization with Keycloak SSO
39mSecure Authorization for Agentic AI in Multi-Domain Environments
01h 06mSupercharge #AgenticAI Projects A DevEx Driven Approach to #CloudNative Scaffolding
27mSimple Control: Managing Multi-Domain MCP Access Via Keycloak IdP With ID-JAG
51mSupercharge Agentic AI Apps: A DevEx Driven Approach to Cloud-Native Scaffolding
27mDesigning an Agentic AI-Ready Secure Authentication and Authorization Platform