Fine Grained Authorisation with Relationship Based Access Control
summary
Argues that relationship-based access control (ReBAC) gives precise, scalable permissions by evaluating paths between entities instead of roles; implements a working demo using Permy (Docker + Postman) that models dog/sheep/bone relationships and toggles access with a make/break friend demo; shows a simple query requiring 8 checks, illustrating CPU cost growth as graph size increases; concludes ReBAC is powerful for document/team scenarios but costly and unnecessary for simple router or login checks.
AI-written summary · about
chapters (13)
- 0:00🚀 Welcome to Build Stuff Conference
- 1:00Introduction to Access Control
- 4:00Role-Based Access Control (RBAC)
- 6:00Attribute-Based Access Control (ABAC)
- 12:00Introduction to Relationship-Based Access Control (ReBAC)
- 15:00Exploring the Structure of ReBAC
- 18:00ReBAC: How It Differs from RBAC & ABAC
- 21:00Complexity of ReBAC – Understanding the Tradeoffs
- 25:00ReBAC Under the Hood – How it Works Technically
- 30:00Real-World Use Cases for ReBAC
- 35:00Practical Demo of ReBAC with Permy
- 40:00Benefits and Challenges of Using ReBAC
- 44:00Conclusion & Key Takeaways
related talks
41mFine Grained Authorisation with Relationship-Based Access Control
49mFine Grained Authorisation with Relationship-Based Access Control
46mInfuse Business Logic into Permissions with ReBAC
46mRebuilding Access Management with ReBAC
45mReBAC at Scale: Let’s Save the p99
44mSecuring RAG Pipelines with Fine Grained Authorization
39mSecuring RAG Pipelines with Fine Grained Authorization
33mFine-Grained Policies RBAC with OpenFGA