Update inference results and and budget (#181)

DSL/Resql/rag-search/POST/get-testing-connection.sql

@@ -0,0 +1,25 @@
+SELECT 
+    id,
+    connection_name,
+    used_budget,
+    monthly_budget,
+    warn_budget_threshold,
+    stop_budget_threshold,
+    environment,
+    connection_status,
+    created_at,
+    llm_platform,
+    llm_model,
+    embedding_platform,
+    embedding_model,
+    CASE 
+        WHEN used_budget IS NULL OR used_budget = 0 OR (used_budget::DECIMAL / monthly_budget::DECIMAL) < (warn_budget_threshold::DECIMAL / 100.0) THEN 'within_budget'
+        WHEN stop_budget_threshold != 0 AND (used_budget::DECIMAL / monthly_budget::DECIMAL) >= (stop_budget_threshold::DECIMAL / 100.0) THEN 'over_budget'
+        WHEN stop_budget_threshold = 0 AND (used_budget::DECIMAL / monthly_budget::DECIMAL) >= 1 THEN 'over_budget'
+        WHEN (used_budget::DECIMAL / monthly_budget::DECIMAL) >= (warn_budget_threshold::DECIMAL / 100.0) THEN 'close_to_exceed'
+        ELSE 'within_budget'
+    END AS budget_status
+FROM llm_connections
+WHERE environment = 'testing'
+ORDER BY created_at DESC
+LIMIT 1;

DSL/Resql/rag-search/POST/store-production-inference-result.sql → DSL/Resql/rag-search/POST/store-inference-result.sql

@@ -7,6 +7,7 @@ INSERT INTO inference_results (
     embedding_scores,
     final_answer,
     environment,
+    llm_connection_id,
     created_at
 ) VALUES (
     :chat_id,
@@ -17,6 +18,7 @@ INSERT INTO inference_results (
     :embedding_scores::JSONB,
     :final_answer,
     :environment,
+    :llm_connection_id,
     :created_at::timestamp with time zone
 ) RETURNING 
     id, 
@@ -28,4 +30,5 @@ INSERT INTO inference_results (
     embedding_scores, 
     final_answer, 
     environment, 
+    llm_connection_id,
     created_at;

DSL/Ruuter.private/rag-search/POST/inference/results/production/store.yml → DSL/Ruuter.public/rag-search/POST/inference/results/store.yml

@@ -29,6 +29,12 @@ declaration:
       - field: final_answer
         type: string
         description: "LLM's final generated answer"
+      - field: environment
+        type: string
+        description: "Environment identifier (e.g., production, testing)"
+      - field: llm_connection_id
+        type: string
+        description: "Connection identifier"
 
 extract_request_data:
   assign:
@@ -39,6 +45,8 @@ extract_request_data:
     ranked_chunks: ${JSON.stringify(incoming.body.ranked_chunks) || null}
     embedding_scores: ${JSON.stringify(incoming.body.embedding_scores) || null}
     final_answer: ${incoming.body.final_answer}
+    environment: ${incoming.body.environment}
+    llm_connection_id: ${incoming.body.llm_connection_id}
     created_at: ${new Date().toISOString()}
   next: validate_required_fields
 
@@ -51,7 +59,7 @@ validate_required_fields:
 store_production_inference_result:
   call: http.post
   args:
-    url: "[#RAG_SEARCH_RESQL]/store-production-inference-result"
+    url: "[#RAG_SEARCH_RESQL]/store-inference-result"
     body:
       chat_id: ${chat_id}
       user_question: ${user_question}
@@ -60,7 +68,8 @@ store_production_inference_result:
       ranked_chunks: ${ranked_chunks}
       embedding_scores: ${embedding_scores}
       final_answer: ${final_answer}
-      environment: "production"
+      environment: ${environment}
+      llm_connection_id: ${llm_connection_id}
       created_at: ${created_at}
   result: store_result
   next: check_status

Dockerfile.llm_orchestration_service

@@ -21,4 +21,4 @@ RUN uv sync --locked
 EXPOSE 8100
 
 # Run the FastAPI app via uvicorn
-CMD ["uv","run","uvicorn", "src.llm_orchestration_service_api:app", "--host", "0.0.0.0", "--port", "8100"]
+CMD ["uv","run","uvicorn", "src.llm_orchestration_service_api:app", "--host", "0.0.0.0", "--port", "8100"]