File size: 12,095 Bytes
a77376b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
"""
LangGraph workflow implementation for the RAG Q&A Agent.
Defines the agent graph with plan, retrieve, answer, and reflect nodes.
"""

from typing import TypedDict, Annotated, Dict, Any, List
from langgraph.graph import StateGraph, END
from rag_pipeline import RAGPipeline
from llm_utils import LLMHandler
from reflection import ReflectionEvaluator
import operator


# Define the agent state
class AgentState(TypedDict):
    """State passed between nodes in the agent workflow."""
    query: str
    plan: str
    needs_retrieval: bool
    retrieved_context: str
    retrieved_chunks: List[Dict[str, Any]]
    answer: str
    reflection: Dict[str, Any]
    final_response: str
    iteration: int


class RAGAgent:
    """LangGraph-based RAG Q&A Agent with reflection."""
    
    def __init__(
        self,
        rag_pipeline: RAGPipeline,
        llm_handler: LLMHandler,
        reflection_evaluator: ReflectionEvaluator,
        max_iterations: int = 2
    ):
        """
        Initialize the RAG agent.
        
        Args:
            rag_pipeline: RAG pipeline for retrieval
            llm_handler: LLM handler for generation
            reflection_evaluator: Reflection evaluator
            max_iterations: Maximum reflection iterations
        """
        self.rag_pipeline = rag_pipeline
        self.llm_handler = llm_handler
        self.reflection_evaluator = reflection_evaluator
        self.max_iterations = max_iterations
        
        # Build the graph
        self.graph = self._build_graph()
        print("βœ“ RAG Agent workflow initialized")
    
    def _build_graph(self):
        """Build the LangGraph workflow."""
        # Create state graph
        workflow = StateGraph(AgentState)
        
        # Add nodes
        workflow.add_node("plan", self.plan_node)
        workflow.add_node("retrieve", self.retrieve_node)
        workflow.add_node("answer", self.answer_node)
        workflow.add_node("reflect", self.reflect_node)
        
        # Define edges
        workflow.set_entry_point("plan")
        
        # Plan -> Retrieve or Answer
        workflow.add_conditional_edges(
            "plan",
            self.should_retrieve,
            {
                True: "retrieve",
                False: "answer"
            }
        )
        
        # Retrieve -> Answer
        workflow.add_edge("retrieve", "answer")
        
        # Answer -> Reflect
        workflow.add_edge("answer", "reflect")
        
        # Reflect -> End or Answer (for regeneration)
        workflow.add_conditional_edges(
            "reflect",
            self.should_regenerate,
            {
                "accept": END,
                "regenerate": "answer",
                "end": END
            }
        )
        
        return workflow.compile()
    
    def plan_node(self, state: AgentState) -> AgentState:
        """
        Planning node: Analyze query and decide if retrieval is needed.
        
        Args:
            state: Current agent state
            
        Returns:
            Updated state with plan
        """
        print("\n" + "="*60)
        print("πŸ“‹ NODE: PLAN")
        print("="*60 + "\n")
        
        query = state["query"]
        print(f"Query: {query}\n")
        
        # Use LLM to analyze query and create a plan
        planning_prompt = f"""Analyze the following user query and determine if it requires retrieving information from a knowledge base.

User Query: "{query}"

Consider:
1. Is this a factual question that would benefit from specific documentation or knowledge?
2. Is this a general question that can be answered without specific context?
3. Does this query ask about specific concepts, technologies, or topics?

Respond in the following format:
NEEDS_RETRIEVAL: [YES/NO]
REASONING: [Brief explanation]
PLAN: [How you will approach answering this query]"""
        
        system_message = "You are a query planning agent. Analyze queries and determine the best approach to answer them."
        
        plan_response = self.llm_handler.generate(
            planning_prompt,
            system_message
        )
        
        # Parse response
        needs_retrieval = "YES" in plan_response.upper().split("NEEDS_RETRIEVAL:")[1].split("\n")[0] if "NEEDS_RETRIEVAL:" in plan_response.upper() else True
        
        print(f"Plan Response:\n{plan_response}\n")
        print(f"Needs Retrieval: {needs_retrieval}")
        
        state["plan"] = plan_response
        state["needs_retrieval"] = needs_retrieval
        state["iteration"] = 0
        
        print("\n" + "="*60 + "\n")
        
        return state
    
    def should_retrieve(self, state: AgentState) -> bool:
        """Conditional edge: Determine if retrieval is needed."""
        return state["needs_retrieval"]
    
    def retrieve_node(self, state: AgentState) -> AgentState:
        """
        Retrieval node: Retrieve relevant context from vector store.
        
        Args:
            state: Current agent state
            
        Returns:
            Updated state with retrieved context
        """
        print("\n" + "="*60)
        print("πŸ” NODE: RETRIEVE")
        print("="*60 + "\n")
        
        query = state["query"]
        
        # Retrieve context
        context, chunks = self.rag_pipeline.retrieve_context(query, top_k=3)
        
        print(f"Retrieved {len(chunks)} relevant chunks\n")
        
        # Display retrieved content preview
        for i, chunk in enumerate(chunks):
            preview = chunk['content'][:150] + "..." if len(chunk['content']) > 150 else chunk['content']
            print(f"Chunk {i+1} Preview: {preview}\n")
        
        state["retrieved_context"] = context
        state["retrieved_chunks"] = chunks
        
        print("="*60 + "\n")
        
        return state
    
    def answer_node(self, state: AgentState) -> AgentState:
        """
        Answer generation node: Generate answer using LLM.
        
        Args:
            state: Current agent state
            
        Returns:
            Updated state with generated answer
        """
        print("\n" + "="*60)
        print("πŸ’¬ NODE: ANSWER")
        print("="*60 + "\n")
        
        query = state["query"]
        iteration = state.get("iteration", 0)
        
        if iteration > 0:
            print(f"[Regeneration attempt {iteration}]\n")
        
        # Check if we have retrieved context
        if state.get("retrieved_context"):
            # Generate answer with context
            context = state["retrieved_context"]
            
            # Check if this is a regeneration with feedback
            if "reflection" in state and iteration > 0:
                feedback = state["reflection"]["reasoning"]
                answer = self._generate_answer_with_feedback(query, context, feedback)
            else:
                answer = self.llm_handler.generate_with_context(
                    query,
                    context,
                    system_message="You are a helpful AI assistant. Answer questions accurately based on the provided context."
                )
        else:
            # Generate answer without context
            answer = self.llm_handler.generate(
                query,
                system_message="You are a helpful AI assistant. Answer questions concisely and accurately."
            )
        
        print(f"Generated Answer:\n{answer}\n")
        
        state["answer"] = answer
        
        print("="*60 + "\n")
        
        return state
    
    def _generate_answer_with_feedback(
        self,
        query: str,
        context: str,
        feedback: str
    ) -> str:
        """
        Generate answer incorporating feedback from reflection.
        
        Args:
            query: User query
            context: Retrieved context
            feedback: Feedback from reflection
            
        Returns:
            Regenerated answer
        """
        prompt = f"""The previous answer was not satisfactory. Here's the feedback:

{feedback}

Now, please generate a better answer to the following question using the context provided.

Context:
{context}

Question: {query}

Provide a comprehensive, accurate, and relevant answer that addresses the feedback."""
        
        system_message = "You are a helpful AI assistant. Learn from feedback and provide improved answers."
        
        return self.llm_handler.generate(prompt, system_message)
    
    def reflect_node(self, state: AgentState) -> AgentState:
        """
        Reflection node: Evaluate answer quality.
        
        Args:
            state: Current agent state
            
        Returns:
            Updated state with reflection results
        """
        query = state["query"]
        answer = state["answer"]
        context = state.get("retrieved_context", "")
        chunks = state.get("retrieved_chunks", [])
        
        # Evaluate answer
        reflection_result = self.reflection_evaluator.evaluate(
            query,
            answer,
            context,
            chunks
        )
        
        state["reflection"] = reflection_result
        
        return state
    
    def should_regenerate(self, state: AgentState) -> str:
        """
        Conditional edge: Determine if answer should be regenerated.
        
        Args:
            state: Current agent state
            
        Returns:
            Next node or END
        """
        reflection = state["reflection"]
        iteration = state.get("iteration", 0)
        
        recommendation = reflection.get("recommendation", "ACCEPT")
        
        # Accept answer if it's good enough or we've hit max iterations
        if recommendation == "ACCEPT" or iteration >= self.max_iterations:
            state["final_response"] = state["answer"]
            return "accept"
        
        # Regenerate if rejected and we haven't hit max iterations
        if recommendation == "REJECT" and iteration < self.max_iterations:
            state["iteration"] = iteration + 1
            print(f"\n⚠️  Answer rejected. Regenerating (iteration {state['iteration']})...\n")
            return "regenerate"
        
        # Otherwise, accept with partial relevance
        state["final_response"] = state["answer"]
        return "end"
    
    def query(self, question: str) -> Dict[str, Any]:
        """
        Process a query through the agent workflow.
        
        Args:
            question: User question
            
        Returns:
            Complete agent response with all state information
        """
        print("\n" + "="*70)
        print(" "*20 + "πŸ€– RAG Q&A AGENT πŸ€–")
        print("="*70 + "\n")
        print(f"User Query: {question}")
        print("="*70)
        
        # Initialize state
        initial_state = AgentState(
            query=question,
            plan="",
            needs_retrieval=True,
            retrieved_context="",
            retrieved_chunks=[],
            answer="",
            reflection={},
            final_response="",
            iteration=0
        )
        
        # Run the graph
        final_state = self.graph.invoke(initial_state)
        
        # Print final result
        print("\n" + "="*70)
        print("βœ… FINAL RESPONSE")
        print("="*70 + "\n")
        print(final_state["final_response"])
        print("\n" + "="*70 + "\n")
        
        return final_state


def create_rag_agent(
    rag_pipeline: RAGPipeline,
    llm_handler: LLMHandler,
    reflection_evaluator: ReflectionEvaluator,
    max_iterations: int = 2
) -> RAGAgent:
    """
    Create and return a RAG agent instance.
    
    Args:
        rag_pipeline: RAG pipeline for retrieval
        llm_handler: LLM handler for generation
        reflection_evaluator: Reflection evaluator
        max_iterations: Maximum reflection iterations
        
    Returns:
        RAGAgent instance
    """
    return RAGAgent(rag_pipeline, llm_handler, reflection_evaluator, max_iterations)