diff --git a/README.md b/README.md
index 5307a51..7e53c01 100644
--- a/README.md
+++ b/README.md
@@ -181,6 +181,7 @@ A curated collection of **Awesome LLM apps built with RAG, AI Agents, Multi-agen
 *   [⛓️ Basic RAG Chain](rag_tutorials/rag_chain/)
 *   [📠 RAG with Database Routing](rag_tutorials/rag_database_routing/)
 *   [🖼️ Vision RAG](rag_tutorials/vision_rag/)
+*   [🕸️ Knowledge Graph RAG with Citations](rag_tutorials/knowledge_graph_rag_citations/)
 
 ### 💾 LLM Apps with Memory Tutorials
 
diff --git a/rag_tutorials/knowledge_graph_rag_citations/Dockerfile b/rag_tutorials/knowledge_graph_rag_citations/Dockerfile
new file mode 100644
index 0000000..db3c120
--- /dev/null
+++ b/rag_tutorials/knowledge_graph_rag_citations/Dockerfile
@@ -0,0 +1,12 @@
+FROM python:3.11-slim
+
+WORKDIR /app
+
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+COPY knowledge_graph_rag.py .
+
+EXPOSE 8501
+
+CMD ["streamlit", "run", "knowledge_graph_rag.py", "--server.address", "0.0.0.0"]
diff --git a/rag_tutorials/knowledge_graph_rag_citations/README.md b/rag_tutorials/knowledge_graph_rag_citations/README.md
new file mode 100644
index 0000000..191e204
--- /dev/null
+++ b/rag_tutorials/knowledge_graph_rag_citations/README.md
@@ -0,0 +1,171 @@
+# 🔍 Knowledge Graph RAG with Verifiable Citations
+
+A Streamlit application demonstrating how **Knowledge Graph-based Retrieval-Augmented Generation (RAG)** provides multi-hop reasoning with fully verifiable source attribution.
+
+## 🎯 What Makes This Different?
+
+Traditional vector-based RAG finds similar text chunks, but struggles with:
+- Questions requiring information from multiple documents
+- Complex reasoning chains
+- Providing verifiable sources for each claim
+
+**Knowledge Graph RAG** solves these by:
+1. **Building a structured graph** of entities and relationships from documents
+2. **Traversing connections** to find related information (multi-hop reasoning)
+3. **Tracking provenance** so every claim links back to its source
+
+## ✨ Features
+
+| Feature | Description |
+|---------|-------------|
+| 🔗 **Multi-hop Reasoning** | Traverse entity relationships to answer complex questions |
+| 📚 **Verifiable Citations** | Every claim includes source document and text |
+| 🧠 **Reasoning Trace** | See exactly how the answer was derived |
+| 🏠 **Fully Local** | Uses Ollama for LLM, Neo4j for graph storage |
+
+## 🚀 Quick Start
+
+### Prerequisites
+
+1. **Ollama** - Local LLM inference
+   ```bash
+   # Install from https://ollama.ai
+   ollama pull llama3.2
+   ```
+
+2. **Neo4j** - Knowledge graph database
+   ```bash
+   # Using Docker
+   docker run -d \
+     --name neo4j \
+     -p 7474:7474 -p 7687:7687 \
+     -e NEO4J_AUTH=neo4j/password \
+     neo4j:latest
+   ```
+
+### Installation
+
+```bash
+# Clone and navigate
+cd knowledge_graph_rag_citations
+
+# Install dependencies
+pip install -r requirements.txt
+
+# Run the app
+streamlit run knowledge_graph_rag.py
+```
+
+## 📖 How It Works
+
+### Step 1: Document → Knowledge Graph
+
+```
+┌─────────────────┐     ┌──────────────────┐     ┌─────────────────┐
+│   Document      │ ──► │  LLM Extraction  │ ──► │ Knowledge Graph │
+│   (Text/PDF)    │     │  (Entities+Rels) │     │    (Neo4j)      │
+└─────────────────┘     └──────────────────┘     └─────────────────┘
+```
+
+The LLM extracts:
+- **Entities**: People, organizations, concepts, technologies
+- **Relationships**: How entities connect (e.g., "works_for", "created", "uses")
+- **Provenance**: Source document and chunk for each extraction
+
+### Step 2: Query → Multi-hop Traversal
+
+```
+┌─────────┐     ┌─────────────┐     ┌─────────────┐     ┌───────────┐
+│  Query  │ ──► │  Find Start │ ──► │  Traverse   │ ──► │  Context  │
+│         │     │   Entities  │     │  Relations  │     │  + Sources│
+└─────────┘     └─────────────┘     └─────────────┘     └───────────┘
+```
+
+### Step 3: Answer → Verified Citations
+
+```
+┌─────────────┐     ┌─────────────┐     ┌──────────────────┐
+│   Context   │ ──► │  Generate   │ ──► │  Answer with     │
+│ + Sources   │     │   Answer    │     │  [1][2] Citations│
+└─────────────┘     └─────────────┘     └──────────────────┘
+                                                │
+                                                ▼
+                                        ┌──────────────────┐
+                                        │ Citation Details │
+                                        │ • Source Doc     │
+                                        │ • Source Text    │
+                                        │ • Reasoning Path │
+                                        └──────────────────┘
+```
+
+## 🖥️ Usage Example
+
+### 1. Add a Document
+
+Paste or select a sample document. The system extracts entities and relationships:
+
+```
+Document: "GraphRAG was developed by Microsoft Research. 
+           Darren Edge led the project..."
+
+Extracted:
+  ├── Entity: GraphRAG (TECHNOLOGY)
+  ├── Entity: Microsoft Research (ORGANIZATION)  
+  ├── Entity: Darren Edge (PERSON)
+  └── Relationship: Darren Edge --[WORKS_FOR]--> Microsoft Research
+```
+
+### 2. Ask a Question
+
+```
+Question: "Who developed GraphRAG and what organization are they from?"
+```
+
+### 3. Get Verified Answer
+
+```
+Answer: GraphRAG was developed by researchers at Microsoft Research [1], 
+        with Darren Edge leading the project [2].
+
+Citations:
+  [1] Source: AI Research Paper
+      Text: "GraphRAG is a technique developed by Microsoft Research..."
+      
+  [2] Source: AI Research Paper  
+      Text: "...introduced by researchers including Darren Edge..."
+```
+
+## 🔧 Configuration
+
+| Setting | Default | Description |
+|---------|---------|-------------|
+| Neo4j URI | `bolt://localhost:7687` | Neo4j connection string |
+| Neo4j User | `neo4j` | Database username |
+| Neo4j Password | - | Database password |
+| LLM Model | `llama3.2` | Ollama model for extraction/generation |
+
+## 🏗️ Architecture
+
+```
+knowledge_graph_rag_citations/
+├── knowledge_graph_rag.py   # Main Streamlit application
+├── requirements.txt         # Python dependencies
+└── README.md               # This file
+```
+
+### Key Components
+
+- **`KnowledgeGraphManager`**: Neo4j interface for graph operations
+- **`extract_entities_with_llm()`**: LLM-based entity/relationship extraction
+- **`generate_answer_with_citations()`**: Multi-hop RAG with provenance tracking
+
+## 🎓 Learn More
+
+This example is inspired by [VeritasGraph](https://github.com/bibinprathap/VeritasGraph), an enterprise-grade framework for:
+- On-premise knowledge graph RAG
+- Visual reasoning traces (Veritas-Scope)
+- LoRA-tuned LLM integration
+
+## 📝 License
+
+MIT License
diff --git a/rag_tutorials/knowledge_graph_rag_citations/docker-compose.yml b/rag_tutorials/knowledge_graph_rag_citations/docker-compose.yml
new file mode 100644
index 0000000..36da736
--- /dev/null
+++ b/rag_tutorials/knowledge_graph_rag_citations/docker-compose.yml
@@ -0,0 +1,58 @@
+version: '3.8'
+
+services:
+  neo4j:
+    image: neo4j:latest
+    container_name: kg-rag-neo4j
+    ports:
+      - "7474:7474"  # Browser
+      - "7687:7687"  # Bolt
+    environment:
+      - NEO4J_AUTH=neo4j/password
+      - NEO4J_PLUGINS='["apoc"]'
+    volumes:
+      - neo4j_data:/data
+    healthcheck:
+      test: ["CMD-SHELL", "wget --no-verbose --tries=1 --spider http://localhost:7474 || exit 1"]
+      interval: 10s
+      timeout: 5s
+      retries: 5
+
+  # Ollama container (optional - can use local installation instead)
+  ollama:
+    image: ollama/ollama:latest
+    container_name: kg-rag-ollama
+    ports:
+      - "11434:11434"
+    volumes:
+      - ollama_data:/root/.ollama
+    # Pull the model on startup
+    entrypoint: ["/bin/sh", "-c"]
+    command:
+      - |
+        /bin/ollama serve &
+        sleep 5
+        /bin/ollama pull llama3.2
+        wait
+
+  streamlit:
+    build:
+      context: .
+      dockerfile: Dockerfile
+    container_name: kg-rag-app
+    ports:
+      - "8501:8501"
+    environment:
+      - NEO4J_URI=bolt://neo4j:7687
+      - NEO4J_USER=neo4j
+      - NEO4J_PASSWORD=password
+      - OLLAMA_HOST=http://ollama:11434
+    depends_on:
+      neo4j:
+        condition: service_healthy
+      ollama:
+        condition: service_started
+
+volumes:
+  neo4j_data:
+  ollama_data:
diff --git a/rag_tutorials/knowledge_graph_rag_citations/knowledge_graph_rag.py b/rag_tutorials/knowledge_graph_rag_citations/knowledge_graph_rag.py
new file mode 100644
index 0000000..64d3e54
--- /dev/null
+++ b/rag_tutorials/knowledge_graph_rag_citations/knowledge_graph_rag.py
@@ -0,0 +1,521 @@
+"""
+Knowledge Graph RAG with Verifiable Citations
+
+A Streamlit app demonstrating how Knowledge Graph-based RAG provides:
+1. Multi-hop reasoning across documents
+2. Verifiable source attribution for every claim
+3. Transparent reasoning traces
+
+This example uses Ollama for local LLM inference and Neo4j for the knowledge graph.
+"""
+
+import streamlit as st
+import ollama
+from ollama import Client as OllamaClient
+from neo4j import GraphDatabase
+from typing import List, Dict, Tuple
+import re
+import os
+from dataclasses import dataclass
+import json
+import hashlib
+
+# Configure Ollama host from environment (for Docker)
+OLLAMA_HOST = os.environ.get('OLLAMA_HOST', 'http://localhost:11434')
+ollama_client = OllamaClient(host=OLLAMA_HOST)
+
+
+# ============================================================================
+# Data Models
+# ============================================================================
+
+@dataclass
+class Entity:
+    """Represents an entity extracted from documents."""
+    id: str
+    name: str
+    entity_type: str
+    description: str
+    source_doc: str
+    source_chunk: str
+
+
+@dataclass
+class Relationship:
+    """Represents a relationship between entities."""
+    source: str
+    target: str
+    relation_type: str
+    description: str
+    source_doc: str
+
+
+@dataclass
+class Citation:
+    """Represents a verifiable citation for a claim."""
+    claim: str
+    source_document: str
+    source_text: str
+    confidence: float
+    reasoning_path: List[str]
+
+
+@dataclass
+class AnswerWithCitations:
+    """Final answer with full attribution."""
+    answer: str
+    citations: List[Citation]
+    reasoning_trace: List[str]
+
+
+# ============================================================================
+# Knowledge Graph Manager
+# ============================================================================
+
+class KnowledgeGraphManager:
+    """Manages the Neo4j knowledge graph for RAG."""
+    
+    def __init__(self, uri: str, user: str, password: str):
+        self.driver = GraphDatabase.driver(uri, auth=(user, password))
+    
+    def close(self):
+        self.driver.close()
+    
+    def clear_graph(self):
+        """Clear all nodes and relationships."""
+        with self.driver.session() as session:
+            session.run("MATCH (n) DETACH DELETE n")
+    
+    def add_entity(self, entity: Entity):
+        """Add an entity to the knowledge graph."""
+        with self.driver.session() as session:
+            session.run(
+                """
+                MERGE (e:Entity {id: $id})
+                SET e.name = $name,
+                    e.type = $entity_type,
+                    e.description = $description,
+                    e.source_doc = $source_doc,
+                    e.source_chunk = $source_chunk
+                """,
+                id=entity.id,
+                name=entity.name,
+                entity_type=entity.entity_type,
+                description=entity.description,
+                source_doc=entity.source_doc,
+                source_chunk=entity.source_chunk
+            )
+    
+    def add_relationship(self, rel: Relationship):
+        """Add a relationship between entities."""
+        with self.driver.session() as session:
+            session.run(
+                """
+                MATCH (a:Entity {name: $source})
+                MATCH (b:Entity {name: $target})
+                MERGE (a)-[r:RELATES_TO {type: $rel_type}]->(b)
+                SET r.description = $description,
+                    r.source_doc = $source_doc
+                """,
+                source=rel.source,
+                target=rel.target,
+                rel_type=rel.relation_type,
+                description=rel.description,
+                source_doc=rel.source_doc
+            )
+    
+    def find_related_entities(self, entity_name: str, hops: int = 2) -> List[Dict]:
+        """Find entities related within N hops, with full provenance."""
+        with self.driver.session() as session:
+            result = session.run(
+                f"""
+                MATCH path = (start:Entity)-[*1..{hops}]-(related:Entity)
+                WHERE toLower(start.name) CONTAINS toLower($name) OR toLower(start.description) CONTAINS toLower($name)
+                RETURN related.name as name,
+                       related.description as description,
+                       related.source_doc as source,
+                       related.source_chunk as chunk,
+                       [r in relationships(path) | r.description] as path_descriptions
+                LIMIT 20
+                """,
+                name=entity_name, hops=hops
+            )
+            return [dict(record) for record in result]
+    
+    def semantic_search(self, query: str) -> List[Dict]:
+        """Search for relevant entities based on query."""
+        with self.driver.session() as session:
+            # Simple text matching (in production, use vector embeddings)
+            result = session.run(
+                """
+                MATCH (e:Entity)
+                WHERE e.name CONTAINS $query 
+                   OR e.description CONTAINS $query
+                RETURN e.name as name,
+                       e.description as description,
+                       e.source_doc as source,
+                       e.source_chunk as chunk,
+                       e.type as type
+                LIMIT 10
+                """,
+                query=query
+            )
+            return [dict(record) for record in result]
+
+
+# ============================================================================
+# LLM-based Entity Extraction
+# ============================================================================
+
+def extract_entities_with_llm(text: str, source_doc: str, model: str = "llama3.2") -> Tuple[List[Entity], List[Relationship]]:
+    """Use LLM to extract entities and relationships from text."""
+    
+    extraction_prompt = f"""Analyze the following text and extract:
+1. KEY ENTITIES (people, organizations, concepts, technologies, events)
+2. RELATIONSHIPS between these entities
+
+For each entity, provide:
+- name: The entity name
+- type: Category (PERSON, ORGANIZATION, CONCEPT, TECHNOLOGY, EVENT, LOCATION)
+- description: Brief description based on the text
+
+For each relationship, provide:
+- source: Source entity name
+- target: Target entity name  
+- type: Relationship type (e.g., WORKS_FOR, CREATED, USES, LOCATED_IN)
+- description: Description of how they relate
+
+TEXT:
+{text}
+
+Respond in JSON format:
+{{
+  "entities": [
+    {{"name": "...", "type": "...", "description": "..."}}
+  ],
+  "relationships": [
+    {{"source": "...", "target": "...", "type": "...", "description": "..."}}
+  ]
+}}
+"""
+    
+    try:
+        response = ollama_client.chat(
+            model=model,
+            messages=[{"role": "user", "content": extraction_prompt}],
+            format="json"
+        )
+        
+        data = json.loads(response['message']['content'])
+        
+        entities = []
+        for e in data.get('entities', []):
+            entity_id = hashlib.md5(f"{e['name']}_{source_doc}".encode()).hexdigest()[:12]
+            entities.append(Entity(
+                id=entity_id,
+                name=e['name'],
+                entity_type=e['type'],
+                description=e['description'],
+                source_doc=source_doc,
+                source_chunk=text[:200] + "..."
+            ))
+        
+        relationships = []
+        for r in data.get('relationships', []):
+            relationships.append(Relationship(
+                source=r['source'],
+                target=r['target'],
+                relation_type=r['type'],
+                description=r['description'],
+                source_doc=source_doc
+            ))
+        
+        return entities, relationships
+    
+    except Exception as e:
+        st.warning(f"Entity extraction error: {e}")
+        return [], []
+
+
+# ============================================================================
+# Multi-hop RAG with Citations
+# ============================================================================
+
+def generate_answer_with_citations(
+    query: str,
+    graph: KnowledgeGraphManager,
+    model: str = "llama3.2"
+) -> AnswerWithCitations:
+    """
+    Generate an answer using multi-hop graph traversal with full citations.
+    
+    This is the core differentiator: every claim is traced back to source documents.
+    """
+    
+    reasoning_trace = []
+    citations = []
+    
+    # Step 1: Initial semantic search
+    reasoning_trace.append(f"🔍 Searching knowledge graph for: '{query}'")
+    initial_results = graph.semantic_search(query)
+    
+    if not initial_results:
+        return AnswerWithCitations(
+            answer="I couldn't find relevant information in the knowledge graph.",
+            citations=[],
+            reasoning_trace=reasoning_trace
+        )
+    
+    reasoning_trace.append(f"📊 Found {len(initial_results)} initial entities")
+    
+    # Step 2: Multi-hop expansion
+    all_context = []
+    for entity in initial_results[:3]:
+        reasoning_trace.append(f"🔗 Expanding from entity: {entity['name']}")
+        related = graph.find_related_entities(entity['name'], hops=2)
+        
+        for rel in related:
+            all_context.append({
+                "entity": rel['name'],
+                "description": rel['description'],
+                "source": rel['source'],
+                "chunk": rel['chunk'],
+                "path": rel.get('path_descriptions', [])
+            })
+            reasoning_trace.append(f"  → Found related: {rel['name']}")
+    
+    # Step 3: Build context with source tracking
+    context_parts = []
+    source_map = {}
+    
+    for i, ctx in enumerate(all_context):
+        source_key = f"[{i+1}]"
+        context_parts.append(f"{source_key} {ctx['entity']}: {ctx['description']}")
+        source_map[source_key] = {
+            "document": ctx['source'],
+            "text": ctx['chunk'],
+            "entity": ctx['entity']
+        }
+    
+    context_text = "\n".join(context_parts)
+    reasoning_trace.append(f"📝 Built context from {len(context_parts)} sources")
+    
+    # Step 4: Generate answer with citation requirements
+    answer_prompt = f"""Based on the following knowledge graph context, answer the question.
+IMPORTANT: For each claim you make, cite the source using [N] notation.
+
+CONTEXT:
+{context_text}
+
+QUESTION: {query}
+
+Provide a comprehensive answer with inline citations [1], [2], etc. for each claim.
+"""
+    
+    try:
+        response = ollama_client.chat(
+            model=model,
+            messages=[{"role": "user", "content": answer_prompt}]
+        )
+        answer = response['message']['content']
+        reasoning_trace.append("✅ Generated answer with citations")
+        
+        # Step 5: Extract and verify citations
+        citation_refs = re.findall(r'\[(\d+)\]', answer)
+        
+        for ref in set(citation_refs):
+            key = f"[{ref}]"
+            if key in source_map:
+                src = source_map[key]
+                citations.append(Citation(
+                    claim=f"Reference {key}",
+                    source_document=src['document'],
+                    source_text=src['text'],
+                    confidence=0.85,
+                    reasoning_path=[f"Entity: {src['entity']}"]
+                ))
+        
+        reasoning_trace.append(f"🔒 Verified {len(citations)} citations")
+        
+        return AnswerWithCitations(
+            answer=answer,
+            citations=citations,
+            reasoning_trace=reasoning_trace
+        )
+        
+    except Exception as e:
+        return AnswerWithCitations(
+            answer=f"Error generating answer: {e}",
+            citations=[],
+            reasoning_trace=reasoning_trace
+        )
+
+
+# ============================================================================
+# Streamlit UI
+# ============================================================================
+
+def main():
+    st.set_page_config(
+        page_title="Knowledge Graph RAG with Citations",
+        page_icon="🔍",
+        layout="wide"
+    )
+    
+    st.title("🔍 Knowledge Graph RAG with Verifiable Citations")
+    st.markdown("""
+    This demo shows how **Knowledge Graph-based RAG** provides:
+    - **Multi-hop reasoning** across connected information
+    - **Verifiable source attribution** for every claim
+    - **Transparent reasoning traces** you can audit
+    
+    Unlike traditional vector RAG, every answer is traceable to its source documents.
+    """)
+    
+    # Sidebar configuration
+    st.sidebar.header("⚙️ Configuration")
+    
+    neo4j_uri = st.sidebar.text_input("Neo4j URI", "bolt://localhost:7687")
+    neo4j_user = st.sidebar.text_input("Neo4j User", "neo4j")
+    neo4j_password = st.sidebar.text_input("Neo4j Password", type="password", value="password")
+    llm_model = st.sidebar.selectbox("LLM Model", ["llama3.2", "mistral", "phi3"])
+    
+    # Initialize session state
+    if 'graph_initialized' not in st.session_state:
+        st.session_state.graph_initialized = False
+        st.session_state.documents = []
+    
+    # Main content
+    tab1, tab2, tab3 = st.tabs(["📄 Add Documents", "❓ Ask Questions", "🔬 View Graph"])
+    
+    with tab1:
+        st.header("Step 1: Build Knowledge Graph from Documents")
+        
+        sample_docs = {
+            "AI Research Paper": """
+            GraphRAG is a technique developed by Microsoft Research that combines knowledge graphs 
+            with retrieval-augmented generation. Unlike traditional RAG which uses vector similarity,
+            GraphRAG builds a structured knowledge graph from documents, enabling multi-hop reasoning.
+            The technique was introduced by researchers including Darren Edge and Ha Trinh.
+            GraphRAG excels at answering complex questions that require connecting information 
+            from multiple sources, such as "What are the relationships between different research projects?"
+            """,
+            "Company Report": """
+            Acme Corp was founded in 2020 by Jane Smith and John Doe in San Francisco.
+            The company develops AI-powered analytics tools for enterprise customers.
+            Their flagship product, DataSense, uses machine learning to analyze business data.
+            Jane Smith previously worked at Google as a senior engineer on the TensorFlow team.
+            John Doe was a co-founder of StartupX, which was acquired by Microsoft in 2019.
+            Acme Corp raised $50 million in Series B funding led by Sequoia Capital.
+            """
+        }
+        
+        doc_choice = st.selectbox("Choose sample document:", list(sample_docs.keys()))
+        doc_text = st.text_area("Or paste your own document:", sample_docs[doc_choice], height=200)
+        doc_name = st.text_input("Document name:", doc_choice)
+        
+        if st.button("🔨 Extract & Add to Knowledge Graph"):
+            with st.spinner("Extracting entities and relationships..."):
+                try:
+                    graph = KnowledgeGraphManager(neo4j_uri, neo4j_user, neo4j_password)
+                    entities, relationships = extract_entities_with_llm(doc_text, doc_name, llm_model)
+                    
+                    for entity in entities:
+                        graph.add_entity(entity)
+                    
+                    for rel in relationships:
+                        graph.add_relationship(rel)
+                    
+                    graph.close()
+                    
+                    st.success(f"✅ Extracted {len(entities)} entities and {len(relationships)} relationships")
+                    
+                    with st.expander("View Extracted Entities"):
+                        for e in entities:
+                            st.write(f"**{e.name}** ({e.entity_type}): {e.description}")
+                    
+                    with st.expander("View Extracted Relationships"):
+                        for r in relationships:
+                            st.write(f"{r.source} --[{r.relation_type}]--> {r.target}: {r.description}")
+                    
+                    st.session_state.graph_initialized = True
+                    st.session_state.documents.append(doc_name)
+                    
+                except Exception as e:
+                    st.error(f"Error: {e}")
+                    st.info("Make sure Neo4j is running and Ollama has the model pulled.")
+    
+    with tab2:
+        st.header("Step 2: Ask Questions with Verifiable Answers")
+        
+        if not st.session_state.graph_initialized:
+            st.warning("⚠️ Please add documents to the knowledge graph first.")
+        else:
+            st.info(f"📚 Knowledge graph contains documents: {', '.join(st.session_state.documents)}")
+        
+        query = st.text_input("Enter your question:", "What are the key concepts in GraphRAG and who developed it?")
+        
+        if st.button("🔍 Ask with Citations"):
+            with st.spinner("Traversing knowledge graph and generating answer..."):
+                try:
+                    graph = KnowledgeGraphManager(neo4j_uri, neo4j_user, neo4j_password)
+                    result = generate_answer_with_citations(query, graph, llm_model)
+                    graph.close()
+                    
+                    # Display reasoning trace
+                    st.subheader("🧠 Reasoning Trace")
+                    for step in result.reasoning_trace:
+                        st.write(step)
+                    
+                    # Display answer
+                    st.subheader("💬 Answer")
+                    st.markdown(result.answer)
+                    
+                    # Display citations
+                    st.subheader("📚 Source Citations")
+                    if result.citations:
+                        for i, citation in enumerate(result.citations):
+                            with st.expander(f"Citation {i+1}: {citation.source_document}"):
+                                st.write(f"**Source Document:** {citation.source_document}")
+                                st.write(f"**Source Text:** {citation.source_text}")
+                                st.write(f"**Confidence:** {citation.confidence:.0%}")
+                                st.write(f"**Reasoning Path:** {' → '.join(citation.reasoning_path)}")
+                    else:
+                        st.info("No specific citations extracted for this answer.")
+                        
+                except Exception as e:
+                    st.error(f"Error: {e}")
+    
+    with tab3:
+        st.header("🔬 Knowledge Graph Visualization")
+        st.info("This tab shows the structure of your knowledge graph.")
+        
+        if st.button("📊 Show Graph Statistics"):
+            try:
+                graph = KnowledgeGraphManager(neo4j_uri, neo4j_user, neo4j_password)
+                with graph.driver.session() as session:
+                    node_count = session.run("MATCH (n) RETURN count(n) as count").single()['count']
+                    rel_count = session.run("MATCH ()-[r]->() RETURN count(r) as count").single()['count']
+                
+                col1, col2 = st.columns(2)
+                col1.metric("Total Entities", node_count)
+                col2.metric("Total Relationships", rel_count)
+                
+                graph.close()
+            except Exception as e:
+                st.error(f"Error connecting to Neo4j: {e}")
+        
+        if st.button("🗑️ Clear Graph"):
+            try:
+                graph = KnowledgeGraphManager(neo4j_uri, neo4j_user, neo4j_password)
+                graph.clear_graph()
+                graph.close()
+                st.session_state.graph_initialized = False
+                st.session_state.documents = []
+                st.success("Graph cleared!")
+            except Exception as e:
+                st.error(f"Error: {e}")
+
+
+if __name__ == "__main__":
+    main()
diff --git a/rag_tutorials/knowledge_graph_rag_citations/requirements.txt b/rag_tutorials/knowledge_graph_rag_citations/requirements.txt
new file mode 100644
index 0000000..4dc44b9
--- /dev/null
+++ b/rag_tutorials/knowledge_graph_rag_citations/requirements.txt
@@ -0,0 +1,3 @@
+streamlit>=1.28.0
+ollama>=0.1.0
+neo4j>=5.0.0