Apply Q1–Q3 decisions: GraphDB/SPARQL, reuse Qdrant, self-host embeddings via TEI

- Q1 → Ontotext GraphDB (SPARQL). docs/08 fully rewritten with RDF schema,
  example SPARQL queries (transitive deps via property paths, time-windowed
  topic counts, cross-pillar joins via nexa:vectorId).
- Q2 → reuse qdrant_scientific with nexa_* collection prefix; docs/09 step 2
  now points there explicitly.
- Q3 → no OpenAI embeddings. Self-host on the docker host. Use TEI
  (HuggingFace text-embeddings-inference) — single Rust binary, ~500 MB image,
  OpenAI-compatible — instead of Ollama, since we only need embeddings.
- docs/09 Phase-3.4 add-on simplified to a single Ontotext compose snippet
  (Neo4j option dropped) plus repo creation curl.
- docs/11 Q3 marked resolved; new Q15 picks the model (bge-m3 vs nomic-embed)
  and adds the open question of whether SAIA already proxies an embedding
  model that would let us skip TEI entirely.
- docs/03 + CLAUDE.md updated with the new decisions so future runs don't
  re-litigate.

docs/03-architecture.md

@@ -21,11 +21,11 @@ A one-page mental model. For details follow the cross-links.
                        │  SAIA    │  │  Qdrant  │
                        │ LiteLLM  │  │  vector  │
                        └──────────┘  └────┬─────┘
-                                          │ optional
+                                          │
                                           ▼
                                      ┌──────────┐
-                                     │  Graph   │  (Neo4j or Ontotext —
-                                     │   DB     │   see open questions)
+                                     │ GraphDB  │  Ontotext, SPARQL
+                                     │  (RDF)   │  Phase 3.4
                                      └──────────┘
                                  │
                                  ▼ writes
@@ -42,8 +42,9 @@ A one-page mental model. For details follow the cross-links.
 | **Memos** | Interface, voice input, webhook source | docker host LXC 104 → `memos.nuclide.systems` |
 | **n8n** | Workflow / logic engine | docker host LXC 104 → `n8n.nuclide.systems` |
 | **SAIA / LiteLLM** | Model gateway, embeddings, classification | docker host LXC 104 → `ai.nuclide.systems` (LiteLLM internal :4000) |
-| **Qdrant** | Vector memory (semantic recall) | docker host LXC 104 (existing `qdrant_scientific` may be reused — see [11](./11-open-questions.md)) |
-| **Graph DB** | Structural memory (deferred) | not yet deployed |
+| **Qdrant** | Vector memory (semantic recall) | docker host LXC 104 — reuse existing `qdrant_scientific` with `nexa_*` collection prefix |
+| **Ontotext GraphDB** | Structural memory via SPARQL (Phase 3.4) | not yet deployed; see [09-deployment](./09-deployment.md#phase-add-on-graph-db-phase-34) |
+| **TEI** (HF text-embeddings-inference) | Self-hosted embeddings (`bge-m3` / TBD — see [11/Q15](./11-open-questions.md)) | docker host LXC 104, CPU only — single-binary embedding server |
 | **Nextcloud** | Tasks, calendar, mail, files | dedicated LXC 105 → `nc.nuclide.systems` |
 | **ntfy** | Push channel for system alerts | docker host → `ntfy.nuclide.systems` |
 | **Backrest** | Backup orchestration | LXC 103 |
@@ -54,7 +55,7 @@ A one-page mental model. For details follow the cross-links.
 ## Two-pillar memory
 
 - **Qdrant** answers *"what is similar / relevant?"* (cosine search over embeddings).
-- **Graph DB** answers *"who, what depends on what, how is it structured?"* (Cypher / SPARQL).
+- **Ontotext GraphDB** answers *"who, what depends on what, how is it structured?"* (SPARQL over RDF).
 
 Both pillars are queried in parallel for `#nexa:ask` and merged before SAIA generates the final answer. See [08 — GraphRAG architecture](./08-graphrag-architecture.md).