genai-example-app-only.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false,
    "id": "I8F3XGz_dyXc"
   },
   "source": [
    "# Neo4j Generative AI Workshop Example Application\n",
    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/neo4j-product-examples/genai-workshop/blob/main/genai-example-app-only.ipynb)\n",
    "\n",
    " __This notebook contains just the demo application for the LLM content generator and assumes you have already run the main workshop notebook `genai-example-app-only.ipynb`__\n",
    "\n",
    "In this workshop, you will learn how to use Neo4j Knowledge Graphs to make Large Language Models (LLMs) useful for more real-world use cases.\n",
    "\n",
    "We walk through an example that uses real-world customer and product data from a fashion, style, and beauty retailer. We show how you can use a knowledge graph to ground an LLM, enabling it to build tailored marketing content personalized to each customer based on their interests and shared purchase histories. We use a pattern called Retrieval-Augmented Generation (RAG) to accomplish this.  Specifically, one that leverages not only vector search but also graph pattern matching and graph machine learning to provide more relevant personalized results to customers.\n",
    "\n",
    "The workshop walks through the end-to-end process, including the below.  This notebook contains just the final demo application for the LLM content generator.\n",
    "- Building the knowledge graph\n",
    "- Vector search & text embedding\n",
    "- Using graph patterns in Cypher to improve semantic search with context\n",
    "- Further augmenting semantic search with knowledge graph inference & ML\n",
    "- Building the LLM chain and demo app for generating content"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false,
    "id": "cmjr1dz8dyXd"
   },
   "source": [
    "## Setup"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Some Logics\n",
    "1. Make a copy of this notebook in Colab by [clicking here](https://colab.research.google.com/github/neo4j-product-examples/genai-workshop/blob/main/genai-workshop.ipynb).\n",
    "2. Run the pip install below to get the necessary dependencies.  this can take a while. Then run the following cell to import relevant libraries\n"
   ],
   "metadata": {
    "collapsed": false,
    "id": "8yxD7Ah0ZACB"
   }
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "outputs": [],
   "source": [
    "%%capture\n",
    "%pip install langchain langchain-openai openai tiktoken python-dotenv gradio neo4j"
   ],
   "metadata": {
    "pycharm": {
     "name": "#%%capture\n"
    },
    "id": "yY1XylsiZACB"
   }
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "id": "7psF1otOdyXe"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from dotenv import load_dotenv\n",
    "import os\n",
    "from langchain_openai import OpenAIEmbeddings, ChatOpenAI\n",
    "from langchain.vectorstores.neo4j_vector import Neo4jVector\n",
    "from langchain.graphs import Neo4jGraph\n",
    "from langchain.prompts import SystemMessagePromptTemplate, HumanMessagePromptTemplate, ChatPromptTemplate\n",
    "from langchain.schema import StrOutputParser\n",
    "from langchain.schema.runnable import RunnableLambda\n",
    "import gradio as gr\n",
    "\n",
    "pd.set_option('display.max_rows', 10)\n",
    "pd.set_option('display.max_colwidth', 500)\n",
    "pd.set_option('display.width', 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false,
    "id": "_ar1ZFhPdyXe"
   },
   "source": [
    "### Setup Credentials and Environment Variables\n",
    "\n",
    "There are two things you need here.\n",
    "1. Start a blank [Neo4j Sandbox](https://sandbox.neo4j.com/). Get your URI and password and plug them in below.  Do not change the Neo4j username.\n",
    "2. Get your OpenAI API key.  You can use [this one](https://docs.google.com/document/d/19Lqjd0MqRs088KUVnd23ZrVU9G0OAg-53U72VrFwwms/edit) if you do not have one already\n",
    "\n",
    "To make this easy, you can write the credentials and env variables directly into the below cell."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "id": "BQ9s0ZWhekd8"
   },
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "# Neo4j\n",
    "NEO4J_URI = 'bolt://34.202.229.218:7687' #change this\n",
    "NEO4J_PASSWORD = 'terminologies-fire-planet' #change this\n",
    "NEO4J_USERNAME = 'neo4j'\n",
    "AURA_DS = False\n",
    "\n",
    "# AI\n",
    "LLM = 'gpt-4'\n",
    "\n",
    "# OpenAI - Required when using OpenAI models\n",
    "os.environ['OPENAI_API_KEY'] = 'sk-...' #change this"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "id": "o-98NuINdyXe"
   },
   "outputs": [],
   "source": [
    "# You can skip this cell if not using a ws.env file - alternative to above\n",
    "from dotenv import load_dotenv\n",
    "import os\n",
    "\n",
    "if os.path.exists('ws-aura.env'):\n",
    "    load_dotenv('ws-aura.env', override=True)\n",
    "\n",
    "    # Neo4j\n",
    "    NEO4J_URI = os.getenv('NEO4J_URI')\n",
    "    NEO4J_USERNAME = os.getenv('NEO4J_USERNAME')\n",
    "    NEO4J_PASSWORD = os.getenv('NEO4J_PASSWORD')\n",
    "    AURA_DS = eval(os.getenv('AURA_DS').title())\n",
    "\n",
    "    # AI\n",
    "    LLM = os.getenv('LLM')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false,
    "id": "90E9HGu4dyXq"
   },
   "source": [
    "## LLM For Generating Grounded Content\n",
    "\n",
    "Let's use an LLM to automatically generate content for targeted marketing campaigns grounded with our knowledge graph using the above tools.\n",
    "Here is a quick example for generating promotional messages, but you can create all sorts of content with this!\n",
    "\n",
    "For our first message, let's consider a scenario where a user recently searched for products, but perhaps didn't commit to a purchase yet. We now want to send a message to promote relevant products."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "outputs": [],
   "source": [
    "from langchain_openai import OpenAIEmbeddings\n",
    "\n",
    "embedding_model = OpenAIEmbeddings()\n",
    "embedding_dimension = 1536"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "id": "JI9LVEdKekeH"
   },
   "outputs": [],
   "source": [
    "# Import relevant libraries\n",
    "from langchain.prompts import SystemMessagePromptTemplate, HumanMessagePromptTemplate, ChatPromptTemplate\n",
    "from langchain_openai import ChatOpenAI\n",
    "from langchain.schema import StrOutputParser\n",
    "\n",
    "#Instantiate LLM\n",
    "llm = ChatOpenAI(temperature=0, model_name=LLM, streaming=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false,
    "id": "8I6JesV0ekeH"
   },
   "source": [
    "### Create Knowledge Graph Stores for Retrieval\n",
    "\n",
    "To ground our content generation, we need to define retrievers to pull information from our knowledge graph.  Let's make two stores:\n",
    "1. Personalized Search Retriever (`kg_personalized_search`): Based on recent customer searches and purchase history, pull relevant products.\n",
    "2. Recommendations retriever (`kg_recommendations_app`): Based on our Graph ML, what else can we recommend to them to pair with the relevant products?\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "id": "WLBBVRXwdyXq"
   },
   "outputs": [],
   "source": [
    "# We will use a mock URL for our sources in the metadata\n",
    "\n",
    "def kg_personalized_search_gen(customer_id):\n",
    "    return Neo4jVector.from_existing_index(\n",
    "        embedding=embedding_model,\n",
    "        url=NEO4J_URI,\n",
    "        username=NEO4J_USERNAME,\n",
    "        password=NEO4J_PASSWORD,\n",
    "        index_name='product_text_embeddings',\n",
    "        retrieval_query=f\"\"\"\n",
    "        WITH node AS product, score AS searchScore\n",
    "\n",
    "        OPTIONAL MATCH(product)<-[:VARIANT_OF]-(:Article)<-[:PURCHASED]-(:Customer)\n",
    "        -[:PURCHASED]->(a:Article)<-[:PURCHASED]-(:Customer {{customerId: '{customer_id}'}})\n",
    "        WITH count(a) AS purchaseScore, product, searchScore\n",
    "        RETURN product.text + '\\nurl: ' + 'https://representative-domain/product/' + product.productCode  AS text,\n",
    "            (1.0+purchaseScore)*searchScore AS score,\n",
    "            {{source: 'https://representative-domain/product/' + product.productCode}} AS metadata\n",
    "        ORDER BY purchaseScore DESC, searchScore DESC LIMIT 5\n",
    "\n",
    "    \"\"\"\n",
    "    )\n",
    "\n",
    "# Use the same personalized recommendations as above but with a smaller limit\n",
    "kg = Neo4jGraph(url=NEO4J_URI, username=NEO4J_USERNAME, password=NEO4J_PASSWORD)\n",
    "def kg_recommendations_app(customer_id, k=30):\n",
    "    res = kg.query(\"\"\"\n",
    "    MATCH(:Customer {customerId:$customerId})-[:PURCHASED]->(:Article)\n",
    "    -[r:CUSTOMERS_ALSO_LIKE]->(:Article)-[:VARIANT_OF]->(product)\n",
    "    RETURN product.text + '\\nurl: ' + 'https://representative-domain/product/' + product.productCode  AS text,\n",
    "        sum(r.score) AS recommenderScore\n",
    "    ORDER BY recommenderScore DESC LIMIT $k\n",
    "    \"\"\", params={'customerId': customer_id, 'k':k})\n",
    "\n",
    "    return \"\\n\\n\".join([d['text'] for d in res])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false,
    "id": "i7sCt8roekeH"
   },
   "source": [
    "### Prompt Engineering\n",
    "\n",
    "Now, let's define our prompts. We will combine two:\n",
    "1. A system prompt which, in this case, tells the LLM how to generate the message\n",
    "2. A human prompt that just wraps the customer search(es)/interest(s)\n",
    "\n",
    "This will allow us to pass the customer interest(s) to the retriever but then also to the LLM for additional context when drafting the message.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "id": "aUAROR6aekeI"
   },
   "outputs": [],
   "source": [
    "general_system_template = '''\n",
    "You are a personal assistant named Sally for a fashion, home, and beauty company called HRM.\n",
    "write an email to {customerName}, one of your customers, to promote and summarize products relevant for them given the current season / time of year: {timeOfYear} .\n",
    "Please only mention the products listed below. Do not come up with or add any new products to the list.\n",
    "Each product comes with an https `url` field. Make sure to provide that https url with descriptive name text in markdown for each product.\n",
    "\n",
    "---\n",
    "# Relevant Products:\n",
    "{searchProds}\n",
    "\n",
    "# Customer May Also Be Interested In the following\n",
    " (pick items from here that pair with the above products well for the current season / time of year: {timeOfYear}.\n",
    " prioritize those higher in the list if possible):\n",
    "{recProds}\n",
    "---\n",
    "'''\n",
    "general_user_template = \"{searchPrompt}\"\n",
    "messages = [\n",
    "    SystemMessagePromptTemplate.from_template(general_system_template),\n",
    "    HumanMessagePromptTemplate.from_template(general_user_template),\n",
    "]\n",
    "prompt = ChatPromptTemplate.from_messages(messages)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false,
    "id": "TgFbaUt6ekeI"
   },
   "source": [
    "### Create a Chain\n",
    "\n",
    "Now let's put a chain together that will leverage the retrievers, prompts, and LLM model. This is where Langchain shines, putting RAG together in a simple way.\n",
    "\n",
    "In addition to the personalized search and recommendations context, we will allow for some other parameters.\n",
    "\n",
    "1. `timeOfYear`: The time of year as a date, season, month, etc. so the LLM can tailor the language appropriately.\n",
    "2. `customerName`: Ordinarily, this can be pulled from the DB, but it has been scrubbed to maintain anonymity, so we will provide our own name here.\n",
    "\n",
    "You can potentially add other creative parameters here to help the LLM write relevant messages.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "id": "nUpih07QdyXr"
   },
   "outputs": [],
   "source": [
    "from langchain.schema.runnable import RunnableLambda\n",
    "\n",
    "# helper function\n",
    "def format_docs(docs):\n",
    "    return \"\\n\\n\".join([d.page_content for d in docs])\n",
    "\n",
    "# LLM chain\n",
    "def chain_gen(customer_id):\n",
    "    return ({'searchProds': (lambda x:x['searchPrompt']) | kg_personalized_search_gen(customer_id).as_retriever(search_kwargs={\"k\": 100}) | format_docs,\n",
    "             'recProds': (lambda x:customer_id) |  RunnableLambda(kg_recommendations_app),\n",
    "             'customerName': lambda x:x['customerName'],\n",
    "             'timeOfYear': lambda x:x['timeOfYear'],\n",
    "             \"searchPrompt\":  lambda x:x['searchPrompt']}\n",
    "            | prompt\n",
    "            | llm\n",
    "            | StrOutputParser())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false,
    "id": "JjbUGH6WekeI"
   },
   "source": [
    "### Example Runs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "outputs": [],
   "source": [
    "# example inputs\n",
    "CUSTOMER_ID = \"daae10780ecd14990ea190a1e9917da33fe96cd8cfa5e80b67b4600171aa77e0\"\n",
    "search_prompt = 'Oversized Sweaters'"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "id": "lkBdqOVjekeI"
   },
   "outputs": [],
   "source": [
    "chain = chain_gen(CUSTOMER_ID)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "outputs": [],
   "source": [
    "#print(chain.invoke({'searchPrompt':search_prompt, 'customerName':'Alex Smith', 'timeOfYear':'Feb, 2024'}))"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "#### Inspecting the Prompt Sent to the LLM\n",
    "In the above run, the LLM should only be using results from our Neo4j database to populate recommendations. Run the below cell to see the final prompt that was sent to the LLM."
   ],
   "metadata": {
    "id": "Vy3fKoC1E0CC"
   }
  },
  {
   "cell_type": "code",
   "source": [
    "def format_final_prompt(x):\n",
    "   return f'''=== Prompt to send to LLM ===\n",
    "   {x.to_string()}\n",
    "   === End Prompt ===\n",
    "   '''\n",
    "\n",
    "def chain_print_prompt(customer_id):\n",
    "    return ({'searchProds': (lambda x:x['searchPrompt']) | kg_personalized_search_gen(customer_id).as_retriever(search_kwargs={\"k\": 100}) | format_docs,\n",
    "             'recProds': (lambda x:customer_id) |  RunnableLambda(kg_recommendations_app),\n",
    "             'customerName': lambda x:x['customerName'],\n",
    "             'timeOfYear': lambda x:x['timeOfYear'],\n",
    "             \"searchPrompt\":  lambda x:x['searchPrompt']}\n",
    "            | prompt\n",
    "            | format_final_prompt\n",
    "            | StrOutputParser())\n",
    "\n",
    "print( chain_print_prompt(CUSTOMER_ID)\\\n",
    "      .invoke({'searchPrompt':search_prompt, 'customerName':'Alex Smith', 'timeOfYear':'Feb, 2024'}))"
   ],
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Z7-yDDUaD6FD",
    "outputId": "686d5f56-2cb8-49bc-8e29-6eb47cb40762"
   },
   "execution_count": 78,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=== Prompt to send to LLM ===\n",
      "   System: \n",
      "You are a personal assistant named Sally for a fashion, home, and beauty company called HRM.\n",
      "write an email to Alex Smith, one of your customers, to promote and summarize products relevant for them given the current season / time of year: Feb, 2024 .\n",
      "Please only mention the products listed below. Do not come up with or add any new products to the list.\n",
      "Each product comes with an https `url` field. Make sure to provide that https url with descriptive name text in markdown for each product.\n",
      "\n",
      "---\n",
      "# Relevant Products:\n",
      "##Product\n",
      "Name: Queen Sweater\n",
      "Type: Sweater\n",
      "Group: Garment Upper body\n",
      "Garment Type: Jersey Basic\n",
      "Description: Top in lightweight sweatshirt fabric with ribbing around the neckline, cuffs and hem.\n",
      "url: https://representative-domain/product/677930\n",
      "\n",
      "##Product\n",
      "Name: Jess oversize LS\n",
      "Type: Top\n",
      "Group: Garment Upper body\n",
      "Garment Type: Jersey Basic\n",
      "Description: Oversized top in soft jersey made from a cotton blend with dropped shoulders and long sleeves.\n",
      "url: https://representative-domain/product/516712\n",
      "\n",
      "##Product\n",
      "Name: Petar Sweater(1)\n",
      "Type: Sweater\n",
      "Group: Garment Upper body\n",
      "Garment Type: Jersey Basic\n",
      "Description: Oversized top in sturdy sweatshirt fabric with dropped shoulders and ribbing around the neckline, cuffs and hem. Soft brushed inside.\n",
      "url: https://representative-domain/product/557247\n",
      "\n",
      "##Product\n",
      "Name: Mother\n",
      "Type: Sweater\n",
      "Group: Garment Upper body\n",
      "Garment Type: Knitwear\n",
      "Description: Long-sleeved jumper in a soft knit containing some mohair with a ribbed turtle neck, dropped shoulders and slits in the sides. Slightly longer at the back.\n",
      "url: https://representative-domain/product/675408\n",
      "\n",
      "##Product\n",
      "Name: Family Crew Ladies\n",
      "Type: Sweater\n",
      "Group: Garment Upper body\n",
      "Garment Type: Special Offers\n",
      "Description: Top in sweatshirt fabric with dropped shoulders, long sleeves and ribbing around the neckline, cuffs and hem. Soft brushed inside.\n",
      "url: https://representative-domain/product/686265\n",
      "\n",
      "# Customer May Also Be Interested In the following\n",
      " (pick items from here that pair with the above products well for the current season / time of year: Feb, 2024.\n",
      " prioritize those higher in the list if possible):\n",
      "##Product\n",
      "Name: DONT USE ROLAND HOOD\n",
      "Type: Hoodie\n",
      "Group: Garment Upper body\n",
      "Garment Type: Jersey Basic\n",
      "Description: Top in sweatshirt fabric with a lined drawstring hood, kangaroo pocket, long raglan sleeves and ribbing at the cuffs and hem.\n",
      "url: https://representative-domain/product/569974\n",
      "\n",
      "##Product\n",
      "Name: PASTRY SWEATER\n",
      "Type: Sweater\n",
      "Group: Garment Upper body\n",
      "Garment Type: Knitwear\n",
      "Description: Jumper in soft, textured-knit cotton with long raglan sleeves and ribbing around the neckline, cuffs and hem.\n",
      "url: https://representative-domain/product/656401\n",
      "\n",
      "##Product\n",
      "Name: Skinny  RW Ankle Milo Zip\n",
      "Type: Trousers\n",
      "Group: Garment Lower body\n",
      "Garment Type: Trousers Denim\n",
      "Description: 5-pocket, ankle-length jeans in washed stretch denim with hard-worn details, a regular waist, zip fly and button, and skinny legs with a zip at the hems. The jeans are made partly from recycled cotton.\n",
      "url: https://representative-domain/product/682848\n",
      "\n",
      "##Product\n",
      "Name: Bubble Bum Bandeau (1)\n",
      "Type: Bikini top\n",
      "Group: Swimwear\n",
      "Garment Type: Swimwear\n",
      "Description: Fully lined bandeau bikini top with padded cups and removable inserts. Detachable ties at the back of the neck, ties at the back, side support and a silicone trim at the top.\n",
      "url: https://representative-domain/product/642498\n",
      "\n",
      "##Product\n",
      "Name: Haven back detail\n",
      "Type: Bra\n",
      "Group: Underwear\n",
      "Garment Type: Under-, Nightwear\n",
      "Description: Push-up bra in lace and mesh with underwired, moulded, padded cups for a larger bust and fuller cleavage. Lace racer back, narrow adjustable shoulder straps, a wide mesh strap at the back and a metal fastener at the front.\n",
      "url: https://representative-domain/product/660519\n",
      "\n",
      "##Product\n",
      "Name: Dixie tee\n",
      "Type: T-shirt\n",
      "Group: Garment Upper body\n",
      "Garment Type: Jersey Fancy\n",
      "Description: Short top in soft cotton jersey with short sleeves. Contrasting colour trims around the neckline and sleeves.\n",
      "url: https://representative-domain/product/598806\n",
      "\n",
      "##Product\n",
      "Name: Eden SP Andes\n",
      "Type: Bra\n",
      "Group: Underwear\n",
      "Garment Type: Under-, Nightwear\n",
      "Description: Super push-up bra in lace with underwired, thickly padded cups to maximise the bust and create a fuller cleavage. Adjustable shoulder straps, a racer back and metal front fastenings.\n",
      "url: https://representative-domain/product/640129\n",
      "\n",
      "##Product\n",
      "Name: Silver lake\n",
      "Type: Sweater\n",
      "Group: Garment Upper body\n",
      "Garment Type: Knitwear\n",
      "Description: Purl-knit jumper in a cotton blend with a slightly wider neckline and 3/4-length sleeves.\n",
      "url: https://representative-domain/product/244267\n",
      "\n",
      "##Product\n",
      "Name: Rylee flatform\n",
      "Type: Heeled sandals\n",
      "Group: Shoes\n",
      "Garment Type: Shoes\n",
      "Description: Sandals with imitation suede straps, an elastic heel strap and wedge heels. Satin insoles and thermoplastic rubber (TPR) soles. Platform front 2 cm, heel 6 cm.\n",
      "url: https://representative-domain/product/606711\n",
      "\n",
      "##Product\n",
      "Name: Leona Push Mirny\n",
      "Type: Bra\n",
      "Group: Underwear\n",
      "Garment Type: Under-, Nightwear\n",
      "Description: Push-up bra in lace and mesh with underwired, moulded, padded cups for a larger bust and fuller cleavage. Lace racer back, narrow adjustable shoulder straps, a wide mesh strap at the back and metal fastener at the front.\n",
      "url: https://representative-domain/product/511924\n",
      "\n",
      "##Product\n",
      "Name: Survivor\n",
      "Type: Blouse\n",
      "Group: Garment Upper body\n",
      "Garment Type: Blouses\n",
      "Description: Straight-cut blouse in a crêpe weave with a collar, concealed buttons down the front and fake flap front pockets. Yoke with a pleat at the back, long sleeves with pleats and buttoned cuffs, and a straight cut hem with slits in the sides.\n",
      "url: https://representative-domain/product/662328\n",
      "\n",
      "##Product\n",
      "Name: Karin headband\n",
      "Type: Hairband\n",
      "Group: Accessories\n",
      "Garment Type: Accessories\n",
      "Description: Wide hairband in cotton jersey with a twisted detail.\n",
      "url: https://representative-domain/product/620425\n",
      "\n",
      "##Product\n",
      "Name: Petar Sweater(1)\n",
      "Type: Sweater\n",
      "Group: Garment Upper body\n",
      "Garment Type: Jersey Basic\n",
      "Description: Oversized top in sturdy sweatshirt fabric with dropped shoulders and ribbing around the neckline, cuffs and hem. Soft brushed inside.\n",
      "url: https://representative-domain/product/557247\n",
      "\n",
      "##Product\n",
      "Name: Rosemary\n",
      "Type: Dress\n",
      "Group: Garment Full body\n",
      "Garment Type: Dresses Ladies\n",
      "Description: Short dress in woven fabric with 3/4-length sleeves with an opening and ties at the cuffs, and a gently rounded hem. Unlined.\n",
      "url: https://representative-domain/product/753724\n",
      "\n",
      "##Product\n",
      "Name: Lead Superskinny\n",
      "Type: Trousers\n",
      "Group: Garment Lower body\n",
      "Garment Type: Trousers\n",
      "Description: Chinos in stretch twill with a zip fly and button, side pockets, welt back pockets and skinny legs.\n",
      "url: https://representative-domain/product/731142\n",
      "\n",
      "##Product\n",
      "Name: Banks\n",
      "Type: Hoodie\n",
      "Group: Garment Upper body\n",
      "Garment Type: Jersey Basic\n",
      "Description: Long-sleeved top in sweatshirt fabric made from a cotton blend with a double-layered hood, gently dropped shoulders and ribbing at the cuffs and hem. Soft brushed inside.\n",
      "url: https://representative-domain/product/752193\n",
      "\n",
      "##Product\n",
      "Name: Malte r-neck\n",
      "Type: Sweater\n",
      "Group: Garment Upper body\n",
      "Garment Type: Knitwear\n",
      "Description: Jumper in soft, patterned, fine-knit cotton with ribbing around the neckline, cuffs and hem.\n",
      "url: https://representative-domain/product/713577\n",
      "\n",
      "##Product\n",
      "Name: Lassy Linnen LS\n",
      "Type: Sweater\n",
      "Group: Garment Upper body\n",
      "Garment Type: Jersey Basic\n",
      "Description: Long-sleeved top in airy linen jersey.\n",
      "url: https://representative-domain/product/531615\n",
      "\n",
      "##Product\n",
      "Name: Burcu Styling Scarf\n",
      "Type: Scarf\n",
      "Group: Accessories\n",
      "Garment Type: Accessories\n",
      "Description: Scarf in soft, patterned satin.\n",
      "url: https://representative-domain/product/772565\n",
      "\n",
      "##Product\n",
      "Name: Girlfriend R.W Trash\n",
      "Type: Trousers\n",
      "Group: Garment Lower body\n",
      "Garment Type: Trousers Denim\n",
      "Description: 5-pocket, ankle-length jeans in washed denim with hard-worn details in a slightly looser fit. Regular waist, zip fly and button, slightly lower crotch and tapered legs with raw-edge hems.\n",
      "url: https://representative-domain/product/724904\n",
      "\n",
      "##Product\n",
      "Name: Veronica dress\n",
      "Type: Dress\n",
      "Group: Garment Full body\n",
      "Garment Type: Dresses Ladies\n",
      "Description: Short dress in a crêpe weave with a V-neck and wrapover front with concealed press-studs. Short sleeves and a seam at the waist with a sewn-in tie belt.\n",
      "url: https://representative-domain/product/710729\n",
      "\n",
      "##Product\n",
      "Name: Brad LW BF Denim TRS\n",
      "Type: Trousers\n",
      "Group: Garment Lower body\n",
      "Garment Type: Trousers\n",
      "Description: 5-pocket, low-rise jeans in washed denim with hard-worn details, a zip fly and button, and slightly wider, tapered legs.\n",
      "url: https://representative-domain/product/615970\n",
      "\n",
      "##Product\n",
      "Name: Derek\n",
      "Type: Dress\n",
      "Group: Garment Full body\n",
      "Garment Type: Dresses Ladies\n",
      "Description: Calf-length dress in woven fabric with a collar, long sleeves and wide cuffs with a slit. Narrow elasticated seam at the waist, a pleated skirt and laser-cut hem. Unlined.\n",
      "url: https://representative-domain/product/706366\n",
      "\n",
      "##Product\n",
      "Name: Gwen Jersey Top\n",
      "Type: Vest top\n",
      "Group: Garment Upper body\n",
      "Garment Type: Dresses Ladies\n",
      "Description: Fitted top in stretch jersey with a slight sheen. V-neck with a lace trim at the top and adjustable spaghetti straps.\n",
      "url: https://representative-domain/product/671852\n",
      "\n",
      "##Product\n",
      "Name: W COLOSSEO BLOUSE EQ\n",
      "Type: Blouse\n",
      "Group: Garment Upper body\n",
      "Garment Type: Unknown\n",
      "Description: Blouse in a cotton weave with a small, frilled collar, concealed buttons down the front, short puff sleeves and a gently rounded hem with slits in the sides. Slightly longer at the back.\n",
      "url: https://representative-domain/product/665648\n",
      "\n",
      "##Product\n",
      "Name: Latte RW slacks\n",
      "Type: Trousers\n",
      "Group: Garment Lower body\n",
      "Garment Type: Trousers\n",
      "Description: Ankle-length cigarette trousers in a stretch weave with a regular waist, hook-and-eye fastening and zip fly. Side pockets, fake welt back pockets and tapered legs.\n",
      "url: https://representative-domain/product/777504\n",
      "\n",
      "##Product\n",
      "Name: BISCUIT\n",
      "Type: Sweater\n",
      "Group: Garment Upper body\n",
      "Garment Type: Knitwear\n",
      "Description: Long-sleeved jumper in a soft, fine knit with ribbing around the neckline, cuffs and hem.\n",
      "url: https://representative-domain/product/657852\n",
      "\n",
      "##Product\n",
      "Name: ROLAND SLIM FIT HOOD\n",
      "Type: Hoodie\n",
      "Group: Garment Upper body\n",
      "Garment Type: Jersey Basic\n",
      "Description: Top in lightweight sweatshirt fabric with ribbed side panels and a kangaroo pocket. Jersey-lined hood with a wrapover front, and ribbing at the cuffs and hem. Brushed inside. Slim Fit.\n",
      "url: https://representative-domain/product/667769\n",
      "\n",
      "##Product\n",
      "Name: BRUNO PIQUE HOOD\n",
      "Type: Hoodie\n",
      "Group: Garment Upper body\n",
      "Garment Type: Jersey Basic\n",
      "Description: Jacket in cotton piqué with a lined, drawstring hood, zip down the front, side pockets and ribbing at the cuffs and hem. Regular fit.\n",
      "url: https://representative-domain/product/622370\n",
      "\n",
      "##Product\n",
      "Name: Fav regular polo(1)\n",
      "Type: Polo shirt\n",
      "Group: Garment Upper body\n",
      "Garment Type: Jersey Fancy\n",
      "Description: Polo shirt in cotton piqué with a ribbed collar, button placket, short sleeves with ribbed trims, and slits in the sides.\n",
      "url: https://representative-domain/product/816759\n",
      "---\n",
      "\n",
      "Human: Oversized Sweaters\n",
      "   === End Prompt ===\n",
      "   \n"
     ]
    }
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false,
    "id": "8G_vdFviekeI"
   },
   "source": [
    "Feel free to experiment and try more!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "outputs": [],
   "source": [
    "#print(chain.invoke({'searchPrompt':\"western boots\", 'customerName':'Alex Smith', 'timeOfYear':'Feb, 2024'}))"
   ],
   "metadata": {
    "id": "qeOts3Q4ZACL"
   }
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false,
    "id": "1IU_gedrekeI"
   },
   "source": [
    "### Demo App\n",
    "Now let’s use the above tools to create a demo app with Gradio.  We will need to make a couple more functions, but otherwise easy to fire up from a Notebook!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "id": "A1F0ve3cekeI"
   },
   "outputs": [],
   "source": [
    "# Create a means to generate and cache chains...so we can quickly try different customer ids\n",
    "personalized_search_chain_cache = dict()\n",
    "def get_chain(customer_id):\n",
    "    if customer_id in personalized_search_chain_cache:\n",
    "        return personalized_search_chain_cache[customer_id]\n",
    "    chain = chain_gen(customer_id)\n",
    "    personalized_search_chain_cache[customer_id] = chain\n",
    "    return chain\n",
    "\n",
    "# create multiple demo examples to try\n",
    "examples = [\n",
    "    [\n",
    "        CUSTOMER_ID,\n",
    "        'March, 2024',\n",
    "        'Alex Smith',\n",
    "        'Oversized Sweaters'\n",
    "    ],\n",
    "    [\n",
    "        '819f4eab1fd76b932fd403ae9f427de8eb9c5b64411d763bb26b5c8c3c30f16f',\n",
    "        'March, 2024',\n",
    "        'Robin Fischer',\n",
    "        'Oversized Sweaters'\n",
    "    ],\n",
    "    [\n",
    "        '44b0898ecce6cc1268dfdb0f91e053db014b973f67e34ed8ae28211410910693',\n",
    "        'March, 2024',\n",
    "        'Chris Johnson',\n",
    "        'Oversized Sweaters'\n",
    "    ],\n",
    "    [\n",
    "        '819f4eab1fd76b932fd403ae9f427de8eb9c5b64411d763bb26b5c8c3c30f16f',\n",
    "        'March, 2024',\n",
    "        'Robin Fischer',\n",
    "        'denim jeans'\n",
    "    ],\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "id": "XsBcFQLlekeI",
    "outputId": "974df8d3-1e34-44e2-e355-3c452f3844c6",
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running on local URL:  http://127.0.0.1:7861\n",
      "\n",
      "Could not create share link. Missing file: /Users/zachblumenfeld/opt/anaconda3/envs/genai-workshop/lib/python3.10/site-packages/gradio/frpc_darwin_arm64_v0.2. \n",
      "\n",
      "Please check your internet connection. This can happen if your antivirus software blocks the download of this file. You can install manually by following these steps: \n",
      "\n",
      "1. Download this file: https://cdn-media.huggingface.co/frpc-gradio-0.2/frpc_darwin_arm64\n",
      "2. Rename the downloaded file to: frpc_darwin_arm64_v0.2\n",
      "3. Move the file to this location: /Users/zachblumenfeld/opt/anaconda3/envs/genai-workshop/lib/python3.10/site-packages/gradio\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": "<IPython.core.display.HTML object>",
      "text/html": "<div><iframe src=\"http://127.0.0.1:7861/\" width=\"100%\" height=\"500\" allow=\"autoplay; camera; microphone; clipboard-read; clipboard-write;\" frameborder=\"0\" allowfullscreen></iframe></div>"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Keyboard interruption in main thread... closing server.\n",
      "Killing tunnel 127.0.0.1:7860 <> None\n",
      "Killing tunnel 127.0.0.1:7861 <> None\n"
     ]
    },
    {
     "data": {
      "text/plain": ""
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import gradio as gr\n",
    "\n",
    "def message_generator(*x):\n",
    "    chain = get_chain(x[0])\n",
    "    return chain.invoke({'searchPrompt':x[3], 'customerName':x[2], 'timeOfYear': x[1]})\n",
    "\n",
    "customer_id = gr.Textbox(value=CUSTOMER_ID, label=\"Customer ID\")\n",
    "time_of_year = gr.Textbox(value=\"March, 2024\", label=\"Time Of Year\")\n",
    "search_prompt_txt = gr.Textbox(value='Oversized Sweaters', label=\"Customer Interests(s)\")\n",
    "customer_name = gr.Textbox(value='Alex Smith', label=\"Customer Name\")\n",
    "message_result = gr.Markdown( label=\"Message\")\n",
    "\n",
    "demo = gr.Interface(fn=message_generator,\n",
    "                    inputs=[customer_id, time_of_year, customer_name, search_prompt_txt],\n",
    "                    outputs=message_result,\n",
    "                    examples=examples,\n",
    "                    title=\"🪄 Message Generator 🥳\")\n",
    "demo.launch(share=True, debug=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [],
   "metadata": {
    "collapsed": false
   }
  }
 ],
 "metadata": {
  "colab": {
   "provenance": [],
   "toc_visible": true
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}