搭建自己的ai客服

最新推荐文章于 2024-08-05 15:57:37 发布

过于真实呢

最新推荐文章于 2024-08-05 15:57:37 发布

阅读量354

点赞数 5

分类专栏： jj的ai之路文章标签：人工智能 flask

本文链接：https://blog.csdn.net/weixin_62115482/article/details/139356781

版权

jj的ai之路专栏收录该内容

32 篇文章 0 订阅

订阅专栏

搭建自己的ai客服

本节介绍的ai客服搭建主要服务于初学者，一套下来希望你对前进的道路更加坚定！那我们开始吧

1.准备工作

1.Agent角色设定

随心设计好Agent的角色，并完成相应的个性化文档供大模型学习

在这里插入图片描述

这里的话建议提前构思好，写好专家文档txt，可以与我不一样，无所谓的。

2.准备好openai的key

本文中调用的是openai的gpt-3.5-turbo，经济实惠

2.搭建flask框架(pycharm)

这里的话代码自己研究研究，可以拓展tools的，不懂问问GPT

1.安装环境

pip install -r requirements.txt

requirements.txt内容：

annotated-types==0.6.0
anyio==3.7.1
blinker==1.7.0
certifi==2023.7.22
charset-normalizer==3.3.2
click==8.1.7
distro==1.8.0
exceptiongroup==1.1.3
flask==3.0.0
functions==0.7.0
h11==0.14.0
httpcore==1.0.2
httpx==0.25.1
idna==3.4
itsdangerous==2.1.2
Jinja2==3.1.2
MarkupSafe==2.1.3
openai==1.2.3
packaging==23.2
pydantic==2.4.2
pydantic-core==2.10.1
requests==2.31.0
sniffio==1.3.0
tqdm==4.66.1
typing-extensions==4.8.0
urllib3==2.1.0
werkzeug==3.0.1

2.app.py源码（框架运行）

import json
import os
import time
from flask import Flask, request, jsonify
import openai
from openai import OpenAI
import functions

from packaging import version


required_version = version.parse("1.1.1")
current_version = version.parse(openai.__version__)
OPENAI_API_KEY = '<填你自己的openai-key>'
#OPENAI_API_BASE ='https://api.novelnetwork.online/v1'
if current_version < required_version:
  raise ValueError(
      f"Error: OpenAI version {openai.__version__} is less than the required version 1.1.1"
  )
else:
  print("OpenAI version is compatible.")

# Flask常规操作
app = Flask(__name__)

# KEY写自己的
client = OpenAI(api_key=OPENAI_API_KEY)#,base_url=OPENAI_API_BASE)

# 加载助手
assistant_id = functions.create_assistant(
    client)  # this function comes from "functions.py"


# 得到对话请求
@app.route('/start', methods=['GET'])
def start_conversation():
  print("Starting a new conversation...")
  thread = client.beta.threads.create()
  print(f"New thread created with ID: {thread.id}")
  return jsonify({"thread_id": thread.id})


# 调用GPT和API产生结果
@app.route('/chat', methods=['POST'])
def chat():
  data = request.json
  thread_id = data.get('thread_id')
  user_input = data.get('message', '')

  if not thread_id:
    print("Error: Missing thread_id")
    return jsonify({"error": "Missing thread_id"}), 400

  print(f"Received message: {user_input} for thread ID: {thread_id}")


  client.beta.threads.messages.create(thread_id=thread_id,
                                      role="user",
                                      content=user_input)

  run = client.beta.threads.runs.create(thread_id=thread_id,
                                        assistant_id=assistant_id)

  # 是否需要调用API
  while True:
    run_status = client.beta.threads.runs.retrieve(thread_id=thread_id,
                                                   run_id=run.id)
    # print(f"Run status: {run_status.status}")
    if run_status.status == 'completed':
      break
    elif run_status.status == 'requires_action':
      # 处理各种API
      for tool_call in run_status.required_action.submit_tool_outputs.tool_calls:
        if tool_call.function.name == "create_lead":
          # Process lead creation
          arguments = json.loads(tool_call.function.arguments)
          output = functions.create_lead(arguments["name"], arguments["phone"],
                                          arguments["wechat"], arguments["address"], arguments["summary"], arguments["intention"])
          client.beta.threads.runs.submit_tool_outputs(thread_id=thread_id,
                                                       run_id=run.id,
                                                       tool_outputs=[{
                                                           "tool_call_id":
                                                           tool_call.id,
                                                           "output":
                                                           json.dumps(output)
                                                       }])
      time.sleep(1)  # Wait for a second before checking again

  # 返回结果
  messages = client.beta.threads.messages.list(thread_id=thread_id)
  response = messages.data[0].content[0].text.value

  print(f"Assistant response: {response}")
  return jsonify({"response": response})


if __name__ == '__main__':
  app.run(host='0.0.0.0', port=8080)

3.functions.py源码

import json
import requests
import os
from openai import OpenAI
import openai
from prompts import assistant_instructions


#注意你复制完可能要加上Bearer这个前缀
AIRTABLE_API_KEY = '<填你自己的airtable-key>'
OPENAI_API_KEY = '<填你自己的openai-key>'
#OPENAI_API_BASE ='https://api.novelnetwork.online/v1'
# Init OpenAI Client
client = OpenAI(api_key=OPENAI_API_KEY)#,base_url=OPENAI_API_BASE)


# 创建一个表单数据
def create_lead(name, phone, wechat,address,summary,intention):
  url = "<替换成自己的在线表单地址>"  # 替换自己的（后续有讲）
  headers = {
      "Authorization": AIRTABLE_API_KEY,
      "Content-Type": "application/json"
  }
  data = {
      "records": [{
          "fields": {
              "Name": name,
              "Phone": phone,
              "Wechat": wechat,
              "Address": address,
              "Summary": summary,
              "Intention": intention
          }
      }]
  }
  response = requests.post(url, headers=headers, json=data)
  if response.status_code == 200:
    print("Lead created successfully.")
    return response.json()
  else:
    print(f"Failed to create lead: {response.text}")

# 创建助手
def create_assistant(client):
  assistant_file_path = 'assistant.json'

  # 如果json不存在就创建一个，如果换本地知识库要删除原来的
  if os.path.exists(assistant_file_path):
    with open(assistant_file_path, 'r') as file:
      assistant_data = json.load(file)
      assistant_id = assistant_data['assistant_id']
      print("Loaded existing assistant ID.")
  else:
    file = client.files.create(file=open("<填写ai要学习的文档地址>", "rb"),
                               purpose='assistants')

    assistant = client.beta.assistants.create(
        # 这里面要调用你写好的指令和接下来会用到的API
        instructions=assistant_instructions,
        model="gpt-3.5-turbo",
        tools=[
            {
                "type": "retrieval"  # This adds the knowledge base as a tool
            },
            {
                "type": "function",  # 写表单
                "function": {
                    "name": "create_lead",
                    "description":
                    "Capture lead details and save to Airtable.",
                    "parameters": {
                        "type": "object",
                        "properties": {
                            "name": {
                                "type": "string",
                                "description": "Name of the lead."
                            },
                            "phone": {
                                "type": "string",
                                "description": "Phone number of the lead."
                            },
                            "wechat": {
                                "type": "string",
                                "description": "Phone number of the lead."
                            },
                            "address": {
                                "type": "string",
                                "description": "Address of the lead."
                            },
                            "summary": {
                                "type": "string",
                                "description": "Summarize students’ questions."
                            },
                            "intention": {
                                "type": "string",
                                "description": "Students’ purchase intention."
                            }
                        },
                        "required": ["name", "phone","wechat","address","summary","intention"]
                    }
                }
            }
        ],
        file_ids=[file.id])

    # 要创建一个新的json
    with open(assistant_file_path, 'w') as file:
      json.dump({'assistant_id': assistant.id}, file)
      print("Created a new assistant and saved the ID.")

    assistant_id = assistant.id

  return assistant_id

4.prompts.py源码

assistant_instructions = """
本助手的角色是金牌销售员，擅长根据你学习到的销售话语与客户进行交流。请你用中文描述学员所咨询问题和答案以及客户的购买意向是否强烈，以便销售团队进行进一步营销，不要发给客户我们已经记录信息了。注意每回答一次学员的问题都尽可能要求他留下详细的信息，因为你的核心目的是销售，只有获取到手机号或者微信号
这两个关键信息后，本助手才需要调用create_lead函数来生成表单，需要向该函数提供学员的姓名，电话，微信号，地址，总结的问题和购买意向这六项参数，如果部分信息未知可以用null来替换；如果客户未提供关键信息则不需要调用。
"""

设置ai的提示词，可自行修改，注意写好prompt的三要素：指令、上下文、角色。

3.运行flask并发布公网

1.运行

在这里插入图片描述

flask框架正常运行的样式，这里没有直接写与前端交互的代码，所以第一个http请求会报404，因为我们这个ai的需求其实是嵌入式的（嵌入网页里头，后面会讲嵌入式的前端怎么处理）

2.发布公网

参考文章网址https://blog.csdn.net/fq157856469/article/details/135271176?ops_request_misc=&request_id=&biz_id=102&utm_term=flask%E6%90%AD%E5%BB%BA%E5%85%AC%E7%BD%91%E6%9C%8D%E5%8A%A1%E5%99%A8&utm_medium=distribute.pc_search_result.none-task-blog-2_allsobaiduweb~default-0-135271176.142^v100pc_search_result_base1&spm=1018.2226.3001.4187

1.注册Cpolar（内网穿透）

由于我们的flask框架运行在局域网，要实现公网的ai嵌入需要内网穿透（这里需要花99块买个二级子域名的权限，舍不着孩子套不着ai）

https://www.cpolar.com

点击官网地址进行注册下载客户端

在这里插入图片描述