How to handle GraphCypherQAChain cypher query issues in better way

[AnupamaJeyashree]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

from flask import Flask, jsonify, request
import os
from langchain_openai import AzureChatOpenAI
from langchain_neo4j import Neo4jGraph, GraphCypherQAChain
from langchain_nomic import NomicEmbeddings
from openai import AzureOpenAI
from dotenv import load_dotenv
import json
# Load environment variables
load_dotenv()

# Initialize Flask app
app = Flask(__name__)

# Initialize Neo4jGraph connection
driver = Neo4jGraph(url=URI, username=USERNAME, password=PASSWORD)

# Initialize Azure Chat for completion
client = AzureChatOpenAI(
    azure_endpoint=AZURE_ENDPOINT,
    api_key=AZURE_API_KEY,
    deployment_name=DEPLOYMENT_NAME,
    api_version=API_VERSION,
    temperature=0.7
)


# Using GraphCypherQAChain
chain = GraphCypherQAChain.from_llm(
    llm=client,
    graph=driver,
    return_direct=True,
    verbose=True,
    allow_dangerous_requests=True
)

@app.route('/search', methods=['POST'])
def search():
    user_query = request.json.get('query')
    filters = request.json.get('filter', {})
    
    if not user_query:
        return jsonify({"error": "Query is required"}), 400

    # Generate embeddings
    embedding = generate_embeddings(user_query)
    if not embedding:
        return jsonify({"error": "Embedding generation failed"}), 500


    # Construct prompt for query generation
    prompt_template = """
    User Query: "{user_query}"
    Filters: {filters}
    Embedding: {embedding}
    Generate a Cypher query that:
    1. Uses the embedding similarity (e.g., `vector.similarity.cosine`) to rank results.
    2. Includes the provided filters.
    3. Give the inputs into the query instead of doing parameterization.

    The Cypher query should look like:
    MATCH (r:Record) WHERE r.Director = '{{Director}}' RETURN r ORDER BY vector.similarity.cosine(r.embedding, {{embedding}}) DESC

    """
    rendered_prompt = prompt_template.format(user_query=user_query, filters=filters, embedding=embedding)

    try:
        # Generate the Cypher query
        generated_results = chain.run(rendered_prompt)
        
        return jsonify({"results": generated_results})
    except Exception as e:
        print("Error:", e)
        return jsonify({"error": str(e)}), 500

Start the Flask app
if __name__ == '__main__':
    app.run(host="0.0.0.0", port=8080, debug=True)

Description

Since embedding also has to be included in the query the query doesn't gets completed and throws error.

System Info

langchain==0.3.13
langchain-community==0.3.13
langchain-core==0.3.28
langchain-neo4j==0.2.0
langchain-nomic==0.1.4
langchain-ollama==0.2.2
langchain-openai==0.2.14
langchain-text-splitters==0.3.4
langgraph==0.2.60
langgraph-checkpoint==2.0.9
langgraph-sdk==0.1.48
langsmith==0.2.7

Windows

Python 3.12.8