Use Case Diagrams¶
Use case diagrams define the system boundary, identify all actors, and map capabilities to roles. Diagrams model the target state — open issues are shown as solved features.
01 — Movie Search¶
User-initiated semantic search through the full AI pipeline. Shows Claude Haiku (intent classification), environment-scoped Qdrant collection (Issue #14 solution), IMDb enrichment, and Claude Sonnet candidate generation.

02 — Authentication¶
Register, login (with rate limiting), token refresh, and logout with full refresh token revocation against the PostgreSQL blacklist.

03 — Chat & Q&A¶
Post-confirmation chat: send message, receive SSE-streamed response, paginated session history, and conversation continuation across sessions.

04 — RAG Ingestion (Target Architecture)¶
Parameterized pipeline (Issue #29): runtime selection of embedding provider (Issue #33 / ADR-0008), vector store (Issue #33), chunking strategy (Issue #31), and environment-scoped collection (Issue #14). Metrics exported to GitHub Actions (Issue #21).

05 — System Roles Map¶
All actors (Guest, Authenticated User, Data Engineer, DevOps/Admin, GitHub Actions CI) and the full capability set each role exercises.
