app.py

import json
import os
import logging
import requests
import openai
from azure.identity import DefaultAzureCredential
from flask import Flask, Response, request, jsonify, send_from_directory
from dotenv import load_dotenv

from backend.auth.auth_utils import get_authenticated_user_details
from backend.history.cosmosdbservice import CosmosConversationClient

load_dotenv()

app = Flask(__name__, static_folder="static")

# Static Files
@app.route("/")
def index():
    return app.send_static_file("index.html")

@app.route("/favicon.ico")
def favicon():
    return app.send_static_file('favicon.ico')

@app.route("/manifest.json")
def manifest():
    return app.send_static_file('manifest.json')

@app.route("/logo192.png")
def logo192():
    return app.send_static_file('logo192.png')

@app.route("/logo512.png")
def logo512():
    return app.send_static_file('logo512.png')

@app.route("/assets/<path:path>")
def assets(path):
    return send_from_directory("static/assets", path)


# ACS Integration Settings
AZURE_SEARCH_SERVICE = os.environ.get("AZURE_SEARCH_SERVICE")
AZURE_SEARCH_INDEX = os.environ.get("AZURE_SEARCH_INDEX")
AZURE_SEARCH_KEY = os.environ.get("AZURE_SEARCH_KEY")
AZURE_SEARCH_USE_SEMANTIC_SEARCH = os.environ.get("AZURE_SEARCH_USE_SEMANTIC_SEARCH", "false")
AZURE_SEARCH_SEMANTIC_SEARCH_CONFIG = os.environ.get("AZURE_SEARCH_SEMANTIC_SEARCH_CONFIG", "default")
AZURE_SEARCH_TOP_K = os.environ.get("AZURE_SEARCH_TOP_K", 5)
AZURE_SEARCH_ENABLE_IN_DOMAIN = os.environ.get("AZURE_SEARCH_ENABLE_IN_DOMAIN", "true")
AZURE_SEARCH_CONTENT_COLUMNS = os.environ.get("AZURE_SEARCH_CONTENT_COLUMNS")
AZURE_SEARCH_FILENAME_COLUMN = os.environ.get("AZURE_SEARCH_FILENAME_COLUMN")
AZURE_SEARCH_TITLE_COLUMN = os.environ.get("AZURE_SEARCH_TITLE_COLUMN")
AZURE_SEARCH_URL_COLUMN = os.environ.get("AZURE_SEARCH_URL_COLUMN")
AZURE_SEARCH_VECTOR_COLUMNS = os.environ.get("AZURE_SEARCH_VECTOR_COLUMNS")
AZURE_SEARCH_QUERY_TYPE = os.environ.get("AZURE_SEARCH_QUERY_TYPE")
AZURE_SEARCH_PERMITTED_GROUPS_COLUMN = os.environ.get("AZURE_SEARCH_PERMITTED_GROUPS_COLUMN")

# AOAI Integration Settings
AZURE_OPENAI_RESOURCE = os.environ.get("AZURE_OPENAI_RESOURCE")
AZURE_OPENAI_MODEL = os.environ.get("AZURE_OPENAI_MODEL")
AZURE_OPENAI_ENDPOINT = os.environ.get("AZURE_OPENAI_ENDPOINT")
AZURE_OPENAI_KEY = os.environ.get("AZURE_OPENAI_KEY")
AZURE_OPENAI_TEMPERATURE = os.environ.get("AZURE_OPENAI_TEMPERATURE", 0)
AZURE_OPENAI_TOP_P = os.environ.get("AZURE_OPENAI_TOP_P", 1.0)
AZURE_OPENAI_MAX_TOKENS = os.environ.get("AZURE_OPENAI_MAX_TOKENS", 1000)
AZURE_OPENAI_STOP_SEQUENCE = os.environ.get("AZURE_OPENAI_STOP_SEQUENCE")
AZURE_OPENAI_SYSTEM_MESSAGE = os.environ.get("AZURE_OPENAI_SYSTEM_MESSAGE", "You are an AI assistant that helps people find information.")
AZURE_OPENAI_PREVIEW_API_VERSION = os.environ.get("AZURE_OPENAI_PREVIEW_API_VERSION", "2023-06-01-preview")
AZURE_OPENAI_STREAM = os.environ.get("AZURE_OPENAI_STREAM", "true")
AZURE_OPENAI_MODEL_NAME = os.environ.get("AZURE_OPENAI_MODEL_NAME", "gpt-35-turbo-16k") # Name of the model, e.g. 'gpt-35-turbo-16k' or 'gpt-4'
AZURE_OPENAI_EMBEDDING_ENDPOINT = os.environ.get("AZURE_OPENAI_EMBEDDING_ENDPOINT")
AZURE_OPENAI_EMBEDDING_KEY = os.environ.get("AZURE_OPENAI_EMBEDDING_KEY")


SHOULD_STREAM = True if AZURE_OPENAI_STREAM.lower() == "true" else False

# CosmosDB Integration Settings
AZURE_COSMOSDB_DATABASE = os.environ.get("AZURE_COSMOSDB_DATABASE")
AZURE_COSMOSDB_ACCOUNT = os.environ.get("AZURE_COSMOSDB_ACCOUNT")
AZURE_COSMOSDB_CONVERSATIONS_CONTAINER = os.environ.get("AZURE_COSMOSDB_CONVERSATIONS_CONTAINER")
AZURE_COSMOSDB_ACCOUNT_KEY = os.environ.get("AZURE_COSMOSDB_ACCOUNT_KEY")

# Initialize a CosmosDB client with AAD auth and containers
cosmos_conversation_client = None
if AZURE_COSMOSDB_DATABASE and AZURE_COSMOSDB_ACCOUNT and AZURE_COSMOSDB_CONVERSATIONS_CONTAINER:
    try :
        cosmos_endpoint = f'https://{AZURE_COSMOSDB_ACCOUNT}.documents.azure.com:443/'

        if not AZURE_COSMOSDB_ACCOUNT_KEY:
            credential = DefaultAzureCredential()
        else:
            credential = AZURE_COSMOSDB_ACCOUNT_KEY

        cosmos_conversation_client = CosmosConversationClient(
            cosmosdb_endpoint=cosmos_endpoint, 
            credential=credential, 
            database_name=AZURE_COSMOSDB_DATABASE,
            container_name=AZURE_COSMOSDB_CONVERSATIONS_CONTAINER
        )
    except Exception as e:
        logging.exception("Exception in CosmosDB initialization", e)
        cosmos_conversation_client = None


def is_chat_model():
    if 'gpt-4' in AZURE_OPENAI_MODEL_NAME.lower() or AZURE_OPENAI_MODEL_NAME.lower() in ['gpt-35-turbo-4k', 'gpt-35-turbo-16k']:
        return True
    return False

def should_use_data():
    if AZURE_SEARCH_SERVICE and AZURE_SEARCH_INDEX and AZURE_SEARCH_KEY:
        return True
    return False


def format_as_ndjson(obj: dict) -> str:
    return json.dumps(obj, ensure_ascii=False) + "\n"

def fetchUserGroups(userToken, nextLink=None):
    # Recursively fetch group membership
    if nextLink:
        endpoint = nextLink
    else:
        endpoint = "https://graph.microsoft.com/v1.0/me/transitiveMemberOf?$select=id"
    
    headers = {
        'Authorization': "bearer " + userToken
    }
    try :
        r = requests.get(endpoint, headers=headers)
        if r.status_code != 200:
            return []
        
        r = r.json()
        if "@odata.nextLink" in r:
            nextLinkData = fetchUserGroups(userToken, r["@odata.nextLink"])
            r['value'].extend(nextLinkData)
        
        return r['value']
    except Exception as e:
        return []


def generateFilterString(userToken):
    # Get list of groups user is a member of
    userGroups = fetchUserGroups(userToken)

    # Construct filter string
    if userGroups:
        group_ids = ", ".join([obj['id'] for obj in userGroups])
        return f"{AZURE_SEARCH_PERMITTED_GROUPS_COLUMN}/any(g:search.in(g, '{group_ids}'))"
    
    return None


def prepare_body_headers_with_data(request):
    request_messages = request.json["messages"]

    # Set query type
    query_type = "simple"
    if AZURE_SEARCH_QUERY_TYPE:
        query_type = AZURE_SEARCH_QUERY_TYPE
    elif AZURE_SEARCH_USE_SEMANTIC_SEARCH.lower() == "true" and AZURE_SEARCH_SEMANTIC_SEARCH_CONFIG:
        query_type = "semantic"

    # Set filter
    filter = None
    userToken = None
    if AZURE_SEARCH_PERMITTED_GROUPS_COLUMN:
        userToken = request.headers.get('X-MS-TOKEN-AAD-ACCESS-TOKEN', "")
        filter = generateFilterString(userToken)

    body = {
        "messages": request_messages,
        "temperature": float(AZURE_OPENAI_TEMPERATURE),
        "max_tokens": int(AZURE_OPENAI_MAX_TOKENS),
        "top_p": float(AZURE_OPENAI_TOP_P),
        "stop": AZURE_OPENAI_STOP_SEQUENCE.split("|") if AZURE_OPENAI_STOP_SEQUENCE else None,
        "stream": SHOULD_STREAM,
        "dataSources": [
            {
                "type": "AzureCognitiveSearch",
                "parameters": {
                    "endpoint": f"https://{AZURE_SEARCH_SERVICE}.search.windows.net",
                    "key": AZURE_SEARCH_KEY,
                    "indexName": AZURE_SEARCH_INDEX,
                    "fieldsMapping": {
                        "contentFields": AZURE_SEARCH_CONTENT_COLUMNS.split("|") if AZURE_SEARCH_CONTENT_COLUMNS else [],
                        "titleField": AZURE_SEARCH_TITLE_COLUMN if AZURE_SEARCH_TITLE_COLUMN else None,
                        "urlField": AZURE_SEARCH_URL_COLUMN if AZURE_SEARCH_URL_COLUMN else None,
                        "filepathField": AZURE_SEARCH_FILENAME_COLUMN if AZURE_SEARCH_FILENAME_COLUMN else None,
                        "vectorFields": AZURE_SEARCH_VECTOR_COLUMNS.split("|") if AZURE_SEARCH_VECTOR_COLUMNS else []
                    },
                    "inScope": True if AZURE_SEARCH_ENABLE_IN_DOMAIN.lower() == "true" else False,
                    "topNDocuments": AZURE_SEARCH_TOP_K,
                    "queryType": query_type,
                    "semanticConfiguration": AZURE_SEARCH_SEMANTIC_SEARCH_CONFIG if AZURE_SEARCH_SEMANTIC_SEARCH_CONFIG else "",
                    "roleInformation": AZURE_OPENAI_SYSTEM_MESSAGE,
                    "embeddingEndpoint": AZURE_OPENAI_EMBEDDING_ENDPOINT,
                    "embeddingKey": AZURE_OPENAI_EMBEDDING_KEY,
                    "filter": filter
                }
            }
        ]
    }

    headers = {
        'Content-Type': 'application/json',
        'api-key': AZURE_OPENAI_KEY,
        "x-ms-useragent": "GitHubSampleWebApp/PublicAPI/2.0.0"
    }

    return body, headers


def stream_with_data(body, headers, endpoint, history_metadata={}):
    s = requests.Session()
    response = {
        "id": "",
        "model": "",
        "created": 0,
        "object": "",
        "choices": [{
            "messages": []
        }],
        'history_metadata': history_metadata
    }
    try:
        with s.post(endpoint, json=body, headers=headers, stream=True) as r:
            for line in r.iter_lines(chunk_size=10):
                if line:
                    lineJson = json.loads(line.lstrip(b'data:').decode('utf-8'))
                    if 'error' in lineJson:
                        yield format_as_ndjson(lineJson)
                    response["id"] = lineJson["id"]
                    response["model"] = lineJson["model"]
                    response["created"] = lineJson["created"]
                    response["object"] = lineJson["object"]

                    role = lineJson["choices"][0]["messages"][0]["delta"].get("role")
                    if role == "tool":
                        response["choices"][0]["messages"].append(lineJson["choices"][0]["messages"][0]["delta"])
                    elif role == "assistant": 
                        response["choices"][0]["messages"].append({
                            "role": "assistant",
                            "content": ""
                        })
                    else:
                        deltaText = lineJson["choices"][0]["messages"][0]["delta"]["content"]
                        if deltaText != "[DONE]":
                            response["choices"][0]["messages"][1]["content"] += deltaText

                    yield format_as_ndjson(response)
    except Exception as e:
        yield format_as_ndjson({"error": str(e)})


def conversation_with_data(request_body):
    body, headers = prepare_body_headers_with_data(request)
    base_url = AZURE_OPENAI_ENDPOINT if AZURE_OPENAI_ENDPOINT else f"https://{AZURE_OPENAI_RESOURCE}.openai.azure.com/"
    endpoint = f"{base_url}openai/deployments/{AZURE_OPENAI_MODEL}/extensions/chat/completions?api-version={AZURE_OPENAI_PREVIEW_API_VERSION}"
    history_metadata = request_body.get("history_metadata", {})

    if not SHOULD_STREAM:
        r = requests.post(endpoint, headers=headers, json=body)
        status_code = r.status_code
        r = r.json()
        r['history_metadata'] = history_metadata

        return Response(format_as_ndjson(r), status=status_code)
    else:
        return Response(stream_with_data(body, headers, endpoint, history_metadata), mimetype='text/event-stream')


def stream_without_data(response, history_metadata={}):
    responseText = ""
    for line in response:
        deltaText = line["choices"][0]["delta"].get('content')
        if deltaText and deltaText != "[DONE]":
            responseText += deltaText

        response_obj = {
            "id": line["id"],
            "model": line["model"],
            "created": line["created"],
            "object": line["object"],
            "choices": [{
                "messages": [{
                    "role": "assistant",
                    "content": responseText
                }]
            }],
            "history_metadata": history_metadata
        }
        yield format_as_ndjson(response_obj)


def conversation_without_data(request_body):
    openai.api_type = "azure"
    openai.api_base = AZURE_OPENAI_ENDPOINT if AZURE_OPENAI_ENDPOINT else f"https://{AZURE_OPENAI_RESOURCE}.openai.azure.com/"
    openai.api_version = "2023-03-15-preview"
    openai.api_key = AZURE_OPENAI_KEY

    request_messages = request_body["messages"]
    messages = [
        {
            "role": "system",
            "content": AZURE_OPENAI_SYSTEM_MESSAGE
        }
    ]

    for message in request_messages:
        messages.append({
            "role": message["role"] ,
            "content": message["content"]
        })

    response = openai.ChatCompletion.create(
        engine=AZURE_OPENAI_MODEL,
        messages = messages,
        temperature=float(AZURE_OPENAI_TEMPERATURE),
        max_tokens=int(AZURE_OPENAI_MAX_TOKENS),
        top_p=float(AZURE_OPENAI_TOP_P),
        stop=AZURE_OPENAI_STOP_SEQUENCE.split("|") if AZURE_OPENAI_STOP_SEQUENCE else None,
        stream=SHOULD_STREAM
    )

    history_metadata = request_body.get("history_metadata", {})

    if not SHOULD_STREAM:
        response_obj = {
            "id": response,
            "model": response.model,
            "created": response.created,
            "object": response.object,
            "choices": [{
                "messages": [{
                    "role": "assistant",
                    "content": response.choices[0].message.content
                }]
            }],
            "history_metadata": history_metadata
        }

        return jsonify(response_obj), 200
    else:
        return Response(stream_without_data(response, history_metadata), mimetype='text/event-stream')


@app.route("/conversation", methods=["GET", "POST"])
def conversation():
    request_body = request.json
    return conversation_internal(request_body)

def conversation_internal(request_body):
    try:
        use_data = should_use_data()
        if use_data:
            return conversation_with_data(request_body)
        else:
            return conversation_without_data(request_body)
    except Exception as e:
        logging.exception("Exception in /conversation")
        return jsonify({"error": str(e)}), 500

## Conversation History API ## 
@app.route("/history/generate", methods=["POST"])
def add_conversation():
    authenticated_user = get_authenticated_user_details(request_headers=request.headers)
    user_id = authenticated_user['user_principal_id']

    ## check request for conversation_id
    conversation_id = request.json.get("conversation_id", None)

    try:
        # make sure cosmos is configured
        if not cosmos_conversation_client:
            raise Exception("CosmosDB is not configured")

        # check for the conversation_id, if the conversation is not set, we will create a new one
        history_metadata = {}
        if not conversation_id:
            title = generate_title(request.json["messages"])
            conversation_dict = cosmos_conversation_client.create_conversation(user_id=user_id, title=title)
            conversation_id = conversation_dict['id']
            history_metadata['title'] = title
            history_metadata['date'] = conversation_dict['createdAt']
            
        ## Format the incoming message object in the "chat/completions" messages format
        ## then write it to the conversation history in cosmos
        messages = request.json["messages"]
        if len(messages) > 0 and messages[-1]['role'] == "user":
            cosmos_conversation_client.create_message(
                conversation_id=conversation_id,
                user_id=user_id,
                input_message=messages[-1]
            )
        else:
            raise Exception("No user message found")
        
        # Submit request to Chat Completions for response
        request_body = request.json
        history_metadata['conversation_id'] = conversation_id
        request_body['history_metadata'] = history_metadata
        return conversation_internal(request_body)
       
    except Exception as e:
        logging.exception("Exception in /history/generate")
        return jsonify({"error": str(e)}), 500


@app.route("/history/update", methods=["POST"])
def update_conversation():
    authenticated_user = get_authenticated_user_details(request_headers=request.headers)
    user_id = authenticated_user['user_principal_id']

    ## check request for conversation_id
    conversation_id = request.json.get("conversation_id", None)

    try:
        # make sure cosmos is configured
        if not cosmos_conversation_client:
            raise Exception("CosmosDB is not configured")

        # check for the conversation_id, if the conversation is not set, we will create a new one
        if not conversation_id:
            raise Exception("No conversation_id found")
            
        ## Format the incoming message object in the "chat/completions" messages format
        ## then write it to the conversation history in cosmos
        messages = request.json["messages"]
        if len(messages) > 0 and messages[-1]['role'] == "assistant":
            if len(messages) > 1 and messages[-2]['role'] == "tool":
                # write the tool message first
                cosmos_conversation_client.create_message(
                    conversation_id=conversation_id,
                    user_id=user_id,
                    input_message=messages[-2]
                )
            # write the assistant message
            cosmos_conversation_client.create_message(
                conversation_id=conversation_id,
                user_id=user_id,
                input_message=messages[-1]
            )
        else:
            raise Exception("No bot messages found")
        
        # Submit request to Chat Completions for response
        response = {'success': True}
        return jsonify(response), 200
       
    except Exception as e:
        logging.exception("Exception in /history/update")
        return jsonify({"error": str(e)}), 500

@app.route("/history/delete", methods=["DELETE"])
def delete_conversation():
    ## get the user id from the request headers
    authenticated_user = get_authenticated_user_details(request_headers=request.headers)
    user_id = authenticated_user['user_principal_id']
    
    ## check request for conversation_id
    conversation_id = request.json.get("conversation_id", None)
    try: 
        if not conversation_id:
            return jsonify({"error": "conversation_id is required"}), 400
        
        ## delete the conversation messages from cosmos first
        deleted_messages = cosmos_conversation_client.delete_messages(conversation_id, user_id)

        ## Now delete the conversation 
        deleted_conversation = cosmos_conversation_client.delete_conversation(user_id, conversation_id)

        return jsonify({"message": "Successfully deleted conversation and messages", "conversation_id": conversation_id}), 200
    except Exception as e:
        logging.exception("Exception in /history/delete")
        return jsonify({"error": str(e)}), 500

@app.route("/history/list", methods=["GET"])
def list_conversations():
    authenticated_user = get_authenticated_user_details(request_headers=request.headers)
    user_id = authenticated_user['user_principal_id']

    ## get the conversations from cosmos
    conversations = cosmos_conversation_client.get_conversations(user_id)
    if not isinstance(conversations, list):
        return jsonify({"error": f"No conversations for {user_id} were found"}), 404

    ## return the conversation ids

    return jsonify(conversations), 200

@app.route("/history/read", methods=["POST"])
def get_conversation():
    authenticated_user = get_authenticated_user_details(request_headers=request.headers)
    user_id = authenticated_user['user_principal_id']

    ## check request for conversation_id
    conversation_id = request.json.get("conversation_id", None)
    
    if not conversation_id:
        return jsonify({"error": "conversation_id is required"}), 400

    ## get the conversation object and the related messages from cosmos
    conversation = cosmos_conversation_client.get_conversation(user_id, conversation_id)
    ## return the conversation id and the messages in the bot frontend format
    if not conversation:
        return jsonify({"error": f"Conversation {conversation_id} was not found. It either does not exist or the logged in user does not have access to it."}), 404
    
    # get the messages for the conversation from cosmos
    conversation_messages = cosmos_conversation_client.get_messages(user_id, conversation_id)

    ## format the messages in the bot frontend format
    messages = [{'id': msg['id'], 'role': msg['role'], 'content': msg['content'], 'createdAt': msg['createdAt']} for msg in conversation_messages]

    return jsonify({"conversation_id": conversation_id, "messages": messages}), 200

@app.route("/history/rename", methods=["POST"])
def rename_conversation():
    authenticated_user = get_authenticated_user_details(request_headers=request.headers)
    user_id = authenticated_user['user_principal_id']

    ## check request for conversation_id
    conversation_id = request.json.get("conversation_id", None)
    
    if not conversation_id:
        return jsonify({"error": "conversation_id is required"}), 400
    
    ## get the conversation from cosmos
    conversation = cosmos_conversation_client.get_conversation(user_id, conversation_id)
    if not conversation:
        return jsonify({"error": f"Conversation {conversation_id} was not found. It either does not exist or the logged in user does not have access to it."}), 404

    ## update the title
    title = request.json.get("title", None)
    if not title:
        return jsonify({"error": "title is required"}), 400
    conversation['title'] = title
    updated_conversation = cosmos_conversation_client.upsert_conversation(conversation)

    return jsonify(updated_conversation), 200

@app.route("/history/delete_all", methods=["DELETE"])
def delete_all_conversations():
    ## get the user id from the request headers
    authenticated_user = get_authenticated_user_details(request_headers=request.headers)
    user_id = authenticated_user['user_principal_id']

    # get conversations for user
    try:
        conversations = cosmos_conversation_client.get_conversations(user_id)
        if not conversations:
            return jsonify({"error": f"No conversations for {user_id} were found"}), 404
        
        # delete each conversation
        for conversation in conversations:
            ## delete the conversation messages from cosmos first
            deleted_messages = cosmos_conversation_client.delete_messages(conversation['id'], user_id)

            ## Now delete the conversation 
            deleted_conversation = cosmos_conversation_client.delete_conversation(user_id, conversation['id'])

        return jsonify({"message": f"Successfully deleted conversation and messages for user {user_id}"}), 200
    
    except Exception as e:
        logging.exception("Exception in /history/delete_all")
        return jsonify({"error": str(e)}), 500
    

@app.route("/history/clear", methods=["POST"])
def clear_messages():
    ## get the user id from the request headers
    authenticated_user = get_authenticated_user_details(request_headers=request.headers)
    user_id = authenticated_user['user_principal_id']
    
    ## check request for conversation_id
    conversation_id = request.json.get("conversation_id", None)
    try: 
        if not conversation_id:
            return jsonify({"error": "conversation_id is required"}), 400
        
        ## delete the conversation messages from cosmos
        deleted_messages = cosmos_conversation_client.delete_messages(conversation_id, user_id)

        return jsonify({"message": "Successfully deleted messages in conversation", "conversation_id": conversation_id}), 200
    except Exception as e:
        logging.exception("Exception in /history/clear_messages")
        return jsonify({"error": str(e)}), 500

@app.route("/history/ensure", methods=["GET"])
def ensure_cosmos():
    if not AZURE_COSMOSDB_ACCOUNT:
        return jsonify({"error": "CosmosDB is not configured"}), 404
    
    if not cosmos_conversation_client or not cosmos_conversation_client.ensure():
        return jsonify({"error": "CosmosDB is not working"}), 500

    return jsonify({"message": "CosmosDB is configured and working"}), 200


def generate_title(conversation_messages):
    ## make sure the messages are sorted by _ts descending
    title_prompt = 'Summarize the conversation so far into a 4-word or less title. Do not use any quotation marks or punctuation. Respond with a json object in the format {{"title": string}}. Do not include any other commentary or description.'

    messages = [{'role': msg['role'], 'content': msg['content']} for msg in conversation_messages]
    messages.append({'role': 'user', 'content': title_prompt})

    try:
        ## Submit prompt to Chat Completions for response
        base_url = AZURE_OPENAI_ENDPOINT if AZURE_OPENAI_ENDPOINT else f"https://{AZURE_OPENAI_RESOURCE}.openai.azure.com/"
        openai.api_type = "azure"
        openai.api_base = base_url
        openai.api_version = "2023-03-15-preview"
        openai.api_key = AZURE_OPENAI_KEY
        completion = openai.ChatCompletion.create(    
            engine=AZURE_OPENAI_MODEL,
            messages=messages,
            temperature=1,
            max_tokens=64 
        )
        title = json.loads(completion['choices'][0]['message']['content'])['title']
        return title
    except Exception as e:
        return messages[-2]['content']

if __name__ == "__main__":
    app.run()