feat: add AI assistant with in-session copilot and standalone chat with RAG

Implements three-phase AI assistant feature:
- Phase 0: RAG infrastructure with pgvector embeddings, Voyage AI integration,
  tree chunking service, and semantic search over team's flow library
- Phase 1: In-session copilot panel during flow navigation with contextual
  AI help, current step awareness, and suggested related flows
- Phase 2: Standalone AI chat page with persistent conversation history,
  pin/delete, and configurable retention policies (account-level)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

backend/app/services/embedding_service.py

@@ -0,0 +1,78 @@
+"""Embedding provider abstraction for RAG.
+
+Uses Voyage AI (voyage-3.5, 1024 dims) as the embedding provider.
+Supports document and query input types for asymmetric search.
+"""
+import logging
+from typing import Optional
+
+from app.core.config import settings
+
+logger = logging.getLogger(__name__)
+
+
+async def get_embedding(
+    text: str,
+    input_type: str = "document",
+) -> Optional[list[float]]:
+    """Get embedding vector for text using Voyage AI.
+
+    Args:
+        text: The text to embed.
+        input_type: "document" for indexing, "query" for search queries.
+
+    Returns:
+        List of floats (1024 dims) or None if embedding service unavailable.
+    """
+    if not settings.VOYAGE_API_KEY:
+        logger.warning("VOYAGE_API_KEY not set — embedding service unavailable")
+        return None
+
+    try:
+        import voyageai
+
+        client = voyageai.AsyncClient(api_key=settings.VOYAGE_API_KEY)
+        result = await client.embed(
+            texts=[text],
+            model=settings.EMBEDDING_MODEL,
+            input_type=input_type,
+        )
+        return result.embeddings[0]
+    except Exception as e:
+        logger.error("Embedding failed: %s", e)
+        return None
+
+
+async def get_embeddings_batch(
+    texts: list[str],
+    input_type: str = "document",
+) -> Optional[list[list[float]]]:
+    """Get embedding vectors for multiple texts in a single API call.
+
+    Args:
+        texts: List of texts to embed.
+        input_type: "document" for indexing, "query" for search queries.
+
+    Returns:
+        List of embedding vectors or None if service unavailable.
+    """
+    if not texts:
+        return []
+
+    if not settings.VOYAGE_API_KEY:
+        logger.warning("VOYAGE_API_KEY not set — embedding service unavailable")
+        return None
+
+    try:
+        import voyageai
+
+        client = voyageai.AsyncClient(api_key=settings.VOYAGE_API_KEY)
+        result = await client.embed(
+            texts=texts,
+            model=settings.EMBEDDING_MODEL,
+            input_type=input_type,
+        )
+        return result.embeddings
+    except Exception as e:
+        logger.error("Batch embedding failed: %s", e)
+        return None