diff --git a/examples/bookstore.py b/examples/bookstore.py
new file mode 100644
index 00000000..b228c940
--- /dev/null
+++ b/examples/bookstore.py
@@ -0,0 +1,116 @@
+# N::Chapter {
+#     chapter_index: I64
+# }
+# 
+# N::SubChapter {
+#     title: String,
+#     content: String
+# }
+# 
+# E::Contains {
+#     From: Chapter,
+#     To: SubChapter,
+#     Properties: {
+#     }
+# }
+# 
+# V::Embedding {
+#     chunk: String
+# }
+# 
+# E::EmbeddingOf {
+#     From: SubChapter,
+#     To: Embedding,
+#     Properties: {
+#         chunk: String
+#     }
+# }
+
+# QUERY loaddocs_rag(chapters: [{ id: I64, subchapters: [{ title: String, content: String, chunks: [{chunk: String, vector: [F64]}]}] }]) =>
+#     FOR {id, subchapters} IN chapters {
+#         chapter_node <- AddN<Chapter>({ chapter_index: id })
+#         FOR {title, content, chunks} IN subchapters {
+#             subchapter_node <- AddN<SubChapter>({ title: title, content: content })
+#             AddE<Contains>::From(chapter_node)::To(subchapter_node)
+#             FOR {chunk, vector} IN chunks {
+#                 vec <- AddV<Embedding>(vector)
+#                 AddE<EmbeddingOf>({chunk: chunk})::From(subchapter_node)::To(vec)
+#             }
+#         }
+#     }
+#     RETURN "Success"
+
+# QUERY searchdocs_rag(query: [F64], k: I32) =>
+#     vecs <- SearchV<Embedding>(query, k)
+#     subchapters <- vecs::In<EmbeddingOf>
+#     RETURN subchapters::{title, content}
+
+# QUERY edge_node() => 
+#     e <- N<Chapter>::OutE<Contains>
+#     RETURN e
+
+# QUERY edge_node_id(id: ID) => 
+#     e <- N<Chapter>::OutE<Contains>(id)
+#     RETURN e
+# 
+
+import helix
+
+db = helix.Db()
+
+class Chapter(db.Node):
+    index: helix.I64
+
+db.index(Chapter.index, unique=True)
+
+class SubChapter(db.Node):
+    title: helix.String
+    content: helix.String
+
+    embedding: SubChapterEmbedding
+
+class SubChapterEmbedding(db.Vector(dimensions=1536, hnsw=helix.cosine)):
+    pass
+
+class Contains(db.Edge[Chapter, SubChapter]):
+    pass
+
+class ArgChapter(helix.Struct):
+    id: helix.I64
+    subchapters: helix.List[ArgSubchapter]
+
+class ArgSubchapter(helix.Struct):
+    title: helix.String
+    content: helix.String
+    chunk: helix.Vector
+
+@db.query
+def loaddocs_rag(chapters: helix.List[ArgChapter]) -> helix.String:
+    for c in chapters:
+        c_node = db.add_node(Chapter(index=c.id))
+
+        for sc in c.subchapters:
+            sc_node = db.add_node(SubChapter(
+                title=sc.title,
+                content=sc.content,
+                embedding=SubChapterEmbedding(sc.chunk)
+            ))
+            
+            db.add_edge(Contains(from=c_node, to=sc_node))
+
+    return "Success"
+
+@db.query
+def searchdocs_rag(query: helix.Vector, k: helix.I32) -> helix.Iterator[helix.Map[helix.String, helix.Value]]:
+    # TODO
+    vecs = db.search_vector(query, k)
+    chapters = vecs.incoming_nodes[Contains]
+    return chapters.map(lambda c: {"index": c.index})
+
+@db.query
+def edge_node() -> helix.Iterator[Contains]:
+    return db.nodes[Chapter].outgoing_edges[Contains]
+
+@db.query
+def edge_node_id(id: helix.Id[Chapter]) -> helix.Iterator[Contains]:
+    return db.nodes[Chapter](id=id).outgoing_edges[Contains]
diff --git a/examples/bookstore.ts b/examples/bookstore.ts
new file mode 100644
index 00000000..26458c9a
--- /dev/null
+++ b/examples/bookstore.ts
@@ -0,0 +1,72 @@
+import hx from "helix";
+
+const schema = hx.schema();
+
+const Chapter = schema.defineNode({
+  index: I64,
+});
+
+schema.index(Chapter.index, { unique: true });
+
+const SubChapter = schema.defineNode({
+  title: hx.String,
+  content: hx.String,
+
+  embedding: SubChapterEmbedding,
+});
+
+const SubChapterEmbedding = schema.defineVector({
+  dimensions: 1536,
+  hnsw: hx.cosine,
+});
+
+const Contains = schema.defineEdge({ from: Chapter, to: SubChapter });
+
+const ArgChapter = hx.Struct({
+  id: hx.I64,
+  subchapters: hx.List(ArgSubchapter),
+});
+
+const ArgSubChapter = hx.Struct({
+  title: hx.String,
+  content: hx.String,
+  chunk: hx.Vector,
+});
+
+const loadDocsRag = schema.query({
+  name: "loaddocs_rag",
+  arguments: [hx.List(ArgChapter)],
+  returns: hx.String,
+}, (db, [chapters]) => {
+  chapters.forEach((c) => {
+    const cNode = db.addNode(Chapter({ index: c.id }));
+
+    c.subchapters.forEach((sc) => {
+      const scNode = db.addNode(SubChapter({
+        title: sc.title,
+        content: sc.content,
+        embedding: SubChapterEmbedding(sc.chunk),
+      }));
+
+      db.addEdge(Contains({ from: cNode, to: scNode }));
+    });
+  });
+
+  return "Success";
+});
+
+const edgeNode = schema.query({
+  name: "edge_node",
+  arguments: [],
+  returns: hx.Iterator(Contains),
+}, (db, []) => {
+  return db.nodes[Chapter].outgoingEdges[Contains];
+});
+
+const edgeNodeId = schema.query({
+  name: "edge_node_id",
+  arguments: [hx.Id(Chapter)],
+  returns: hx.Iterator(Contains),
+}, (db, [id]) => {
+  return db.nodes[Chapter]({ id }).outgoingEdges[Contains];
+});