diff --git a/.github/workflows/push.yml b/.github/workflows/push.yml
index 21d08f8..8ab90d9 100644
--- a/.github/workflows/push.yml
+++ b/.github/workflows/push.yml
@@ -14,8 +14,8 @@ env:
   BOT_CURSOR: ${{ vars.BOT_CURSOR }}
   DYNAMODB_TABLE_NAME: ${{ vars.DYNAMODB_TABLE_NAME }}
   KNOWLEDGE_BASE_ID: ${{ vars.KNOWLEDGE_BASE_ID }}
-  MODEL_ID_IMAGE: ${{ vars.MODEL_ID_IMAGE }}
   MODEL_ID_TEXT: ${{ vars.MODEL_ID_TEXT }}
+  PERSONAL_MESSAGE: ${{ vars.PERSONAL_MESSAGE }}
   SYSTEM_MESSAGE: ${{ vars.SYSTEM_MESSAGE }}
 
   AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
@@ -56,10 +56,10 @@ jobs:
           echo "BOT_CURSOR=${BOT_CURSOR}" >> .env
           echo "DYNAMODB_TABLE_NAME=${DYNAMODB_TABLE_NAME}" >> .env
           echo "KNOWLEDGE_BASE_ID=${KNOWLEDGE_BASE_ID}" >> .env
-          echo "MODEL_ID_IMAGE=${MODEL_ID_IMAGE}" >> .env
           echo "MODEL_ID_TEXT=${MODEL_ID_TEXT}" >> .env
           echo "SLACK_BOT_TOKEN=${SLACK_BOT_TOKEN}" >> .env
           echo "SLACK_SIGNING_SECRET=${SLACK_SIGNING_SECRET}" >> .env
+          echo "PERSONAL_MESSAGE=${PERSONAL_MESSAGE}" >> .env
           echo "SYSTEM_MESSAGE=${SYSTEM_MESSAGE}" >> .env
 
       - name: Deploy to AWS Lambda 🚀
diff --git a/handler.py b/handler.py
index 565da86..912cbd6 100644
--- a/handler.py
+++ b/handler.py
@@ -39,6 +39,9 @@
 ALLOWED_CHANNEL_IDS = os.environ.get("ALLOWED_CHANNEL_IDS", "None")
 
 # Set up System messages
+PERSONAL_MESSAGE = os.environ.get(
+    "PERSONAL_MESSAGE", "당신은 친절하고 전문적인 AI 비서 입니다."
+)
 SYSTEM_MESSAGE = os.environ.get("SYSTEM_MESSAGE", "None")
 
 MAX_LEN_SLACK = int(os.environ.get("MAX_LEN_SLACK", 3000))
@@ -304,16 +307,14 @@ def conversation(say: Say, thread_ts, query, channel, client_msg_id):
     latest_ts = result["ts"]
 
     prompts = []
-    prompts.append(
-        "Human: You are a advisor AI system, and provides answers to questions by using fact based and statistical information when possible."
-    )
-    prompts.append(
-        "If you don't know the answer, just say that you don't know, don't try to make up an answer."
-    )
+    prompts.append("Human: {}".format(PERSONAL_MESSAGE))
+    prompts.append("답변을 모르면 모른다고 하세요. 답을 지어내려고 하지 마세요.")
 
     if SYSTEM_MESSAGE != "None":
         prompts.append(SYSTEM_MESSAGE)
 
+    prompts.append("<question> 태그로 감싸진 질문에 답변을 제공하세요.")
+
     try:
         # Get the knowledge base contexts
         if KNOWLEDGE_BASE_ID != "None":
@@ -322,7 +323,7 @@ def conversation(say: Say, thread_ts, query, channel, client_msg_id):
             contexts = invoke_knowledge_base(query)
 
             prompts.append(
-                "Use the following pieces of information to provide a concise answer to the question enclosed in <question> tags."
+                "<context> 에 정보가 제공 되면, 해당 정보를 사용하여 답변해 주세요."
             )
             prompts.append("<context>")
             prompts.append("\n\n".join(contexts))
@@ -334,9 +335,12 @@ def conversation(say: Say, thread_ts, query, channel, client_msg_id):
 
                 contexts = conversations_replies(channel, thread_ts, client_msg_id)
 
-                prompts.append("<context>")
+                prompts.append(
+                    "<history> 에 정보가 제공 되면, 대화 기록을 참고하여 답변해 주세요."
+                )
+                prompts.append("<history>")
                 prompts.append("\n\n".join(contexts))
-                prompts.append("</context>")
+                prompts.append("</history>")
 
         # Add the question to the prompts
         prompts.append("")
@@ -345,7 +349,6 @@ def conversation(say: Say, thread_ts, query, channel, client_msg_id):
         prompts.append("</question>")
         prompts.append("")
 
-        # prompts.append("The response should be specific and use statistics or numbers when possible.")
         prompts.append("Assistant:")
 
         # Combine the prompts