- RAG & SPIN Specialist improvements

config/tasks/IDENTIFICATION_DETECTION_TASK/1.0.0.yaml

@@ -16,10 +16,7 @@ task_description: >
   
   {custom_description}
 expected_output: >
-  - Personal Identification information such as name, email, phone number, job title, and any additional information that
-  may prove to be interesting in the current or future conversations.
-  - Company information such as company name, industry, size, company website, ...
-  {custom_expected_output}
+
 metadata:
   author: "Josako"
   date_added: "2025-01-08"

config/tasks/IDENTIFICATION_QUESTIONS_TASK/1.0.0.yaml

@@ -11,12 +11,7 @@ task_description: >
   
   {custom_description}
 expected_output: >
-  - Personal Identification information such as name, email, phone number, job title, and any additional information that
-    may prove to be interesting in the current or future conversations.
-  - Company information such as company name, industry, size, company website, ...
-    {custom_expected_output}
-  - Top {nr_of_questions} questions to ask in order to complete identification.
-  {custom_expected_output}
+
 metadata:
   author: "Josako"
   date_added: "2025-01-08"

config/tasks/RAG_CONSOLIDATION_TASK/1.0.0.yaml

@@ -19,7 +19,6 @@ task_description: >
   %%%{additional_questions}%%%
 
 expected_output: >
-  One consolidated communication towards the end user.
   {custom_expected_output}
 metadata:
   author: "Josako"

config/tasks/RAG_TASK/1.0.0.yaml

@@ -14,9 +14,7 @@ task_description: >
   Query:
     {query}
 expected_output: >
-  - Answer
-  - A list of sources used in generating the answer, citations
-  - An indication (True or False) if there's insufficient information to give an answer.
+
 metadata:
   author: "Josako"
   date_added: "2025-01-08"

config/tasks/SPIN_DETECT_TASK/1.0.0.yaml

@@ -10,13 +10,7 @@ task_description: >
   Latest reply:
     {query}
 expected_output: >
-  The SPIN analysis, comprised of:
-  - Situation information: a description of the customer's current context / situation.
-  - Problem information: a description of the customer's problems uncovering it's challenges and pain points.
-  - Implication information: implications of situation / identified problems, i.e. of the consequences of those problems.
-  - Need-payoff information: Customer's needs, helping customers realize value of solutions.
-  - Additional info: Information that does not fit in the above SPIN-categories, but that can be commercially 
-  interesting, including if provided: {custom_expected_output}
+
 metadata:
   author: "Josako"
   date_added: "2025-01-08"

config/tasks/SPIN_QUESTIONS_TASK/1.0.0.yaml

@@ -12,17 +12,7 @@ task_description: >
   Latest reply:
     {query}
 expected_output: >
-  The SPIN analysis, comprised of:
-   - Situation information: a description of the customer's current context / situation.
-   - Problem information: a description of the customer's problems uncovering it's challenges and pain points.
-   - Implication information: implications of situation / identified problems, i.e. of the consequences of those problems.
-   - Need-payoff information: Customer's needs, helping customers realize value of solutions.
-   - Additional info: Information that does not fit in the above SPIN-categories, but that can be commercially interesting.
-  The SPIN questions:
-  - At max {nr_of_questions} questions to complete the SPIN-context of the customer, as a markdown list without '```'.
-  Potential Customer Indication:
-  - An indication if - given the current SPIN - this could be a good customer (True) or not (False).
-  {custom_expected_output}
+
 metadata:
   author: "Josako"
   date_added: "2025-01-08"

eveai_chat_workers/outputs/rag/rag_v1_0.py

@@ -4,6 +4,6 @@ from pydantic import BaseModel, Field
 
 
 class RAGOutput(BaseModel):
-    answer: Optional[str] = Field(None, description="Answer to the questions asked")
-    citations: Optional[List[str]] = Field(None, description="A list of sources used in generating the answer")
-    insufficient_info: Optional[bool] = Field(None, description="An indication if there's insufficient information to answer")
+    answer: str = Field(None, description="Answer to the questions asked")
+    insufficient_info: bool = Field(None, description="An indication if there's insufficient information to answer")
+

eveai_chat_workers/retrievers/retriever_typing.py

@@ -1,4 +1,4 @@
-from typing import Dict, Any
+from typing import Dict, Any, Optional
 
 from flask import current_app
 from pydantic import BaseModel, Field, model_validator
@@ -11,6 +11,7 @@ class RetrieverMetadata(BaseModel):
     document_id: int = Field(..., description="ID of the source document")
     version_id: int = Field(..., description="Version ID of the source document")
     document_name: str = Field(..., description="Name of the source document")
+    url: Optional[str] = Field(..., description="URL of the source document")
     user_metadata: Dict[str, Any] = Field(
         default_factory=dict,  # This will use an empty dict if None is provided
         description="User-defined metadata"

eveai_chat_workers/retrievers/standard_rag.py

@@ -97,6 +97,7 @@ class StandardRAGRetriever(BaseRetriever):
             query_obj = (
                 db.session.query(
                     db_class,
+                    DocumentVersion.url,
                     (1 - db_class.embedding.cosine_distance(query_embedding)).label('similarity')
                 )
                 .join(DocumentVersion, db_class.doc_vers_id == DocumentVersion.id)
@@ -116,7 +117,7 @@ class StandardRAGRetriever(BaseRetriever):
 
             # Transform results into standard format
             processed_results = []
-            for doc, similarity in results:
+            for doc, url, similarity in results:
                 # Parse user_metadata to ensure it's a dictionary
                 user_metadata = self._parse_metadata(doc.document_version.user_metadata)
                 processed_results.append(
@@ -128,6 +129,7 @@ class StandardRAGRetriever(BaseRetriever):
                             document_id=doc.document_version.doc_id,
                             version_id=doc.document_version.id,
                             document_name=doc.document_version.document.name,
+                            url=url or "",
                             user_metadata=user_metadata,
                         )
                     )

eveai_chat_workers/specialists/RAG_SPECIALIST/1_0.py

@@ -65,7 +65,6 @@ class SpecialistExecutor(CrewAIBaseSpecialistExecutor):
             "language": arguments.language,
             "query": arguments.query,
             "context": formatted_context,
-            "citations": citations,
             "history": self.formatted_history,
             "name": self.specialist.configuration.get('name', ''),
             "company": self.specialist.configuration.get('company', ''),

eveai_chat_workers/specialists/SPIN_SPECIALIST/1_0.py

@@ -124,7 +124,6 @@ class SpecialistExecutor(CrewAIBaseSpecialistExecutor):
             "language": arguments.language,
             "query": arguments.query,
             "context": formatted_context,
-            "citations": citations,
             "history": self.formatted_history,
             "historic_spin": json.dumps(self.latest_spin, indent=2),
             "historic_lead_info": json.dumps(self.latest_lead_info, indent=2),
@@ -217,7 +216,9 @@ class SPINFlow(EveAICrewAIFlow[SPINFlowState]):
     async def execute_rag(self):
         inputs = self.state.input.model_dump()
         try:
+            current_app.logger.debug("In execute_rag")
             crew_output = await self.rag_crew.kickoff_async(inputs=inputs)
+            current_app.logger.debug(f"Crew execution ended with output:\n{crew_output}")
             self.specialist_executor.log_tuning("RAG Crew Output", crew_output.model_dump())
             output_pydantic = crew_output.pydantic
             if not output_pydantic:
@@ -276,13 +277,16 @@ class SPINFlow(EveAICrewAIFlow[SPINFlowState]):
         if self.state.spin:
             additional_questions = additional_questions + self.state.spin.questions
         inputs["additional_questions"] = additional_questions
+        current_app.logger.debug(f"Prepared Answers: \n{inputs['prepared_answers']}")
+        current_app.logger.debug(f"Additional Questions: \n{additional_questions}")
         try:
             crew_output = await self.rag_consolidation_crew.kickoff_async(inputs=inputs)
+            current_app.logger.debug(f"Consolidation output after crew execution:\n{crew_output}")
             self.specialist_executor.log_tuning("RAG Consolidation Crew Output", crew_output.model_dump())
             output_pydantic = crew_output.pydantic
             if not output_pydantic:
                 raw_json = json.loads(crew_output.raw)
-                output_pydantic = LeadInfoOutput.model_validate(raw_json)
+                output_pydantic = RAGOutput.model_validate(raw_json)
             self.state.final_output = output_pydantic
             return crew_output
         except Exception as e:

eveai_chat_workers/specialists/crewai_base_specialist.py

@@ -314,7 +314,10 @@ class CrewAIBaseSpecialistExecutor(BaseSpecialistExecutor):
         ])
 
         # Return document_ids for citations
-        citations = [ctx.metadata.document_id for ctx in unique_contexts]
+        citations = [{"document_id": ctx.metadata.document_id,
+                     "document_version_id": ctx.metadata.version_id,
+                      "url": ctx.metadata.url}
+                     for ctx in unique_contexts]
 
         self.log_tuning("Context Retrieval Results",
                         {"Formatted Context": formatted_context,
@@ -345,6 +348,7 @@ class CrewAIBaseSpecialistExecutor(BaseSpecialistExecutor):
 
         modified_result = {
             "detailed_query": detailed_query,
+            "citations": citations,
         }
         final_result = result.model_copy(update=modified_result)