수정 - AI 및 fastAPI 초기 구조 셋팅

ai-server/backend/ai/embedding.py

@@ -1,9 +1,12 @@
 from langchain_ibm import WatsonxEmbeddings
 from backend.config.settings import settings
+from langchain_ollama import OllamaEmbeddings
 
 watson_embedding = WatsonxEmbeddings(
     model_id="ibm/granite-embedding-278m-multilingual",
     url=f"{settings.watsonx_url}",
     api_key=f"{settings.watsonx_api_key}",
     project_id=f"{settings.watsonx_project_id}",
-)
+)
+
+ollama_embedding = OllamaEmbeddings(model="nomic-embed-text-v2-moe")

ai-server/backend/ai/local_llm.py

@@ -0,0 +1,6 @@
+from langchain_ollama import ChatOllama
+
+# 로컬 LLM
+qwen_llm = ChatOllama(model="qwen3.5:4b", temperature=0)
+exaone_llm = ChatOllama(model="exaone3.5:2.4b", temperature=0)
+gemma_llm = ChatOllama(model="gemma4:e2b")

ai-server/backend/main.py

@@ -1,44 +1,29 @@
 from fastapi import FastAPI
 from starlette.staticfiles import StaticFiles
 from contextlib import asynccontextmanager
-
-from backend.routers.stock_router import router as stock_router
+from backend.routers.npc_router import router as npc_router
+from backend.routers.skill_router import router as skill_router
 
 from backend.repository.db_init import Base, SessionLocal, engine
 
-# from backend.routers.evalu_router import router as evalu_router
-# from backend.routers.assistant_router import router as assistant_router
-# from backend.routers.call_router import router as call_router
-# from backend.repository.seed import seed_customers
-
-@asynccontextmanager
-async def lifespan(app: FastAPI):
-    print("서버 시작")
-
-    # Base 에 등록된 모든 모델에 테이블 자동 생성
-    Base.metadata.create_all(bind=engine)
-    print("[DB] 테이블 생성 완료 (또는 이미 존재)")
-
-    yield
-    print("서버 종료")
+# @asynccontextmanager
+# async def lifespan(app: FastAPI):
+#     print("서버 시작")
+#
+#     # Base 에 등록된 모든 모델에 테이블 자동 생성
+#     Base.metadata.create_all(bind=engine)
+#     print("[DB] 테이블 생성 완료 (또는 이미 존재)")
+#
+#     yield
+#     print("서버 종료")
 
 # app = FastAPI()
 # app = FastAPI(title="상담 LLM", version="1.0", lifespan=lifespan)
-app = FastAPI(title="$ Stock AI", version="1.0", lifespan=lifespan, description="""
-    #### $주식 AI 분석 FastAPI + LangGraph + yfinance + ChromaDB
-    
-    ### 주요 기능
-    | 번호 | 기능 | 엔드포인트 |
-    | --- | --- | --- |
-    |1 | 기업 종합 분석 | 'POST /api/stock/analysis'
-    |2 | 투자 의견 조회 | 'GET /api/stock/opinion/{analysis_id}'
-    """,
-)
+app = FastAPI(title="AI npc", version="1.0", description="던전 입구를 지키는 문지기 npc",)
 
 # static 폴더 지정
 # app.mount("/static", StaticFiles(directory="backend/static"), name="static")
 
 # 라우터 등록
-app.include_router(stock_router)
-# app.include_router(assistant_router)
-# app.include_router(evalu_router)
+app.include_router(npc_router)
+app.include_router(skill_router)

ai-server/backend/routers/npc_router.py

@@ -0,0 +1,13 @@
+from fastapi import APIRouter, Request, UploadFile
+from backend.services.llm_service import question_and_answer
+from backend.schemas.basic_schema import QuestionRequest
+from fastapi.responses import StreamingResponse
+
+router = APIRouter(prefix="/npc")
+
+# http://127.0.0.1:8000/npc/question
+@router.post("/question")
+async def question(req: QuestionRequest):
+    answer = question_and_answer(req.question)
+
+    return {"message" : answer}

ai-server/backend/routers/skill_router.py

@@ -0,0 +1,13 @@
+from fastapi import APIRouter, Request, UploadFile
+from backend.services.llm_service import question_and_answer
+from backend.schemas.basic_schema import QuestionRequest
+from fastapi.responses import StreamingResponse
+
+router = APIRouter(prefix="/npc")
+
+# http://127.0.0.1:8000/npc/question
+@router.post("/question")
+async def question(req: QuestionRequest):
+    answer = question_and_answer(req.question)
+
+    return {"message" : answer}

ai-server/backend/schemas/basic_schema.py

@@ -0,0 +1,4 @@
+from pydantic import BaseModel
+
+class QuestionRequest(BaseModel):
+    question : str

ai-server/backend/services/llm_service.py

@@ -0,0 +1,8 @@
+from backend.ai.llm import watson_llm
+
+# LLM 모델 통신
+# 데이터베이스 통신
+
+def question_and_answer(question):
+    response = watson_llm.invoke(question)
+    return response.content