feat: Qdrant external retriever

.github/workflows/pr-e2e-tests.yaml

@@ -45,6 +45,10 @@ jobs:
         ports:
           - 7687:7687
           - 7474:7474
+      qdrant:
+        image: qdrant/qdrant
+        ports:
+          - 6333:6333
 
     steps:
       - name: Install graphviz package

.github/workflows/scheduled-e2e-tests.yaml

@@ -52,6 +52,10 @@ jobs:
         credentials:
           username: ${{ secrets.DOCKERHUB_USERNAME }}
           password: ${{ secrets.DOCKERHUB_TOKEN }}
+      qdrant:
+        image: qdrant/qdrant
+        ports:
+          - 6333:6333
 
     steps:
       - name: Install graphviz package

CHANGELOG.md

@@ -35,6 +35,7 @@
 - Added support for Cohere LLM and embeddings - added optional dependency to `cohere`.
 - Added support for Anthropic LLM - added optional dependency to `anthropic`.
 - Added support for MistralAI LLM - added optional dependency to `mistralai`.
+- Added support for Qdrant - added optional dependency to `qdrant-client`.
 
 ### Fixed
 - Resolved import issue with the Vertex AI Embeddings class.

docs/source/api.rst

@@ -163,6 +163,12 @@ PineconeNeo4jRetriever
 .. autoclass:: neo4j_graphrag.retrievers.external.pinecone.pinecone.PineconeNeo4jRetriever
     :members: search
 
+QdrantNeo4jRetriever
+====================
+
+.. autoclass:: neo4j_graphrag.retrievers.external.qdrant.qdrant.QdrantNeo4jRetriever
+    :members: search
+
 
 ********
 Embedder

docs/source/user_guide_rag.rst

@@ -327,6 +327,8 @@ We provide implementations for the following retrievers:
      - Use this retriever when vectors are saved in a Weaviate vector database
    * - :ref:`PineconeNeo4jRetriever <pinecone-neo4j-retriever-user-guide>`
      - Use this retriever when vectors are saved in a Pinecone vector database
+   * - :ref:`QdrantNeo4jRetriever <qdrant-neo4j-retriever-user-guide>`
+     - Use this retriever when vectors are saved in a Qdrant vector database
 
 Retrievers all expose a `search` method that we will discuss in the next sections.
 
@@ -672,6 +674,35 @@ Pinecone Retrievers
 
 Also see :ref:`pineconeneo4jretriever`.
 
+.. _qdrant-neo4j-retriever-user-guide:
+
+Qdrant Retrievers
+-----------------
+
+.. note::
+
+    In order to import this retriever, the Qdrant Python client must be installed:
+    `pip install qdrant-client`
+
+
+.. code:: python
+
+    from qdrant_client import QdrantClient
+    from neo4j_graphrag.retrievers import QdrantNeo4jRetriever
+
+    client = QdrantClient(...)  # construct the Qdrant client instance
+
+    retriever = QdrantNeo4jRetriever(
+        driver=driver,
+        client=client,
+        collection_name="my-collection",
+        id_property_external="neo4j_id",    # The payload field that contains identifier to a corresponding Neo4j node id property
+        id_property_neo4j="id",
+        embedder=embedder,
+    )
+
+See :ref:`qdrantneo4jretriever`.
+
 
 Other Retrievers
 ===================

examples/qdrant/README.md

@@ -0,0 +1,31 @@
+### Start services locally
+
+Run the following command to spin up Neo4j and Qdrant containers.
+
+```bash
+docker compose -f tests/e2e/docker-compose.yml up
+```
+
+### Write data (once)
+
+Run this from the project root to write data to both Neo4J and Qdrant.
+
+```bash
+poetry run python tests/e2e/qdrant_e2e/populate_dbs.py
+```
+
+### Install Qdrant client
+
+```bash
+pip install qdrant-client
+```
+
+### Search
+
+```bash
+# search by vector
+poetry run python -m examples.qdrant.vector_search
+
+# search by text, with embeddings generated locally
+poetry run python -m examples.qdrant.text_search
+```

examples/qdrant/text_search.py

@@ -0,0 +1,27 @@
+from langchain_huggingface.embeddings import HuggingFaceEmbeddings
+from neo4j import GraphDatabase
+from neo4j_graphrag.retrievers import QdrantNeo4jRetriever
+from qdrant_client import QdrantClient
+
+NEO4J_URL = "neo4j://localhost:7687"
+NEO4J_AUTH = ("neo4j", "password")
+
+
+def main() -> None:
+    with GraphDatabase.driver(NEO4J_URL, auth=NEO4J_AUTH) as neo4j_driver:
+        embedder = HuggingFaceEmbeddings(model_name="all-MiniLM-L6-v2")
+        retriever = QdrantNeo4jRetriever(
+            driver=neo4j_driver,
+            client=QdrantClient(url="http://localhost:6333"),
+            collection_name="Jeopardy",
+            id_property_external="neo4j_id",
+            id_property_neo4j="id",
+            embedder=embedder,  # type: ignore
+        )
+
+        res = retriever.search(query_text="biology", top_k=2)
+        print(res)
+
+
+if __name__ == "__main__":
+    main()

examples/qdrant/vector_search.py

@@ -0,0 +1,25 @@
+from neo4j import GraphDatabase
+from neo4j_graphrag.retrievers import QdrantNeo4jRetriever
+from qdrant_client import QdrantClient
+
+from examples.embedding_biology import EMBEDDING_BIOLOGY
+
+NEO4J_URL = "neo4j://localhost:7687"
+NEO4J_AUTH = ("neo4j", "password")
+
+
+def main() -> None:
+    with GraphDatabase.driver(NEO4J_URL, auth=NEO4J_AUTH) as neo4j_driver:
+        retriever = QdrantNeo4jRetriever(
+            driver=neo4j_driver,
+            client=QdrantClient(url="http://localhost:6333"),
+            collection_name="Jeopardy",
+            id_property_external="neo4j_id",
+            id_property_neo4j="id",
+        )
+        res = retriever.search(query_vector=EMBEDDING_BIOLOGY, top_k=2)
+        print(res)
+
+
+if __name__ == "__main__":
+    main()