Chatbot de questions-réponses sur les règles de cricket IPL avec RAG et Google Gemini API

Name: Chatbot de questions-réponses sur les règles de cricket IPL avec RAG et Google Gemini API
Rating: 4.5 (10 reviews)
Author: Sidd

Avancé

Ceci est unEngineering, Multimodal AIworkflow d'automatisation du domainecontenant 24 nœuds.Utilise principalement des nœuds comme HttpRequest, ManualTrigger, Agent, ChatTrigger, LmChatGoogleGemini. Chatbot de questions-réponses sur les règles de cricket IPL basé sur RAG et l'API Google Gemini

Prérequis

•Peut nécessiter les informations d'identification d'authentification de l'API cible
•Clé API Google Gemini

Nœuds utilisés (24)

EmbeddingsGoogleGemini

DocumentDefaultDataLoader

TextSplitterRecursiveCharacterTextSplitter

Catégorie

Ingénierie

IA Multimodale

Aperçu du workflow

Visualisation des connexions entre les nœuds, avec support du zoom et du déplacement

À la réception d'un message de chat

Agent IA

Mémoire simple

Vector Store simple

Modèle de chat Google Gemini

Embeddings Google Gemini

Vector Store simple1

Embeddings Google Gemini1

Chargeur de données par défaut

Séparateur de texte récursif

Requête HTTP

Au clic sur 'Exécuter le workflow'

React Flow

Exporter le workflow

Copiez la configuration JSON suivante dans n8n pour importer et utiliser ce workflow

{
  "id": "CkgF5zRqCL4BS6I5",
  "meta": {
    "instanceId": "5c50f3d58b333c0490a31213f0ec76116e02346dcdd9088649ba9dd6fbe45ca1",
    "templateCredsSetupCompleted": true
  },
  "name": "IPL Cricket Rules Q&A Chat Bot using RAG and Google Gemini API",
  "tags": [],
  "nodes": [
    {
      "id": "4c32f558-efff-4eff-b714-202c7419a96c",
      "name": "À la réception d'un message de chat",
      "type": "@n8n/n8n-nodes-langchain.chatTrigger",
      "position": [
        -1216,
        192
      ],
      "webhookId": "4df707a8-70c8-4fab-a970-a97ce7d7594f",
      "parameters": {
        "options": {}
      },
      "typeVersion": 1.1
    },
    {
      "id": "352186bb-07d1-4d7d-9f0f-b57e0880fc11",
      "name": "Agent IA",
      "type": "@n8n/n8n-nodes-langchain.agent",
      "position": [
        -1008,
        64
      ],
      "parameters": {
        "options": {
          "systemMessage": "You are a cricket expert. \n\nYou are tasked with answering questions on ipl cricket queries. Information should only be referred to and provided if it is provided explicitly in the data base to you. Your goal is to provide accurate information based on this information.\n\nIf information is not provided to you explicitly or if you can not answer the question using the provided information, say \"Sorry I donot know\""
        }
      },
      "typeVersion": 2.1
    },
    {
      "id": "15f7fbdc-ab77-4007-9a8e-8ddbe881d984",
      "name": "Mémoire simple",
      "type": "@n8n/n8n-nodes-langchain.memoryBufferWindow",
      "position": [
        -784,
        336
      ],
      "parameters": {
        "contextWindowLength": 20
      },
      "typeVersion": 1.3
    },
    {
      "id": "dc61d50a-fdd8-4a21-974f-33aa8aab5c0a",
      "name": "Vector Store simple",
      "type": "@n8n/n8n-nodes-langchain.vectorStoreInMemory",
      "position": [
        -720,
        176
      ],
      "parameters": {
        "mode": "retrieve-as-tool",
        "topK": 10,
        "memoryKey": {
          "__rl": true,
          "mode": "list",
          "value": "vector_store_key"
        },
        "toolDescription": "This is a repository of ipl cricket rules and international cricket rules"
      },
      "typeVersion": 1.3
    },
    {
      "id": "69f8782c-c5d2-4693-bc00-a2ab58c61e08",
      "name": "Modèle de chat Google Gemini",
      "type": "@n8n/n8n-nodes-langchain.lmChatGoogleGemini",
      "position": [
        -944,
        336
      ],
      "parameters": {
        "options": {
          "topP": 0.3
        }
      },
      "credentials": {
        "googlePalmApi": {
          "id": "3f4CCF4BMZnEfG6y",
          "name": "Google Gemini(PaLM) Api account"
        }
      },
      "typeVersion": 1
    },
    {
      "id": "33d9a2a4-6f13-4cbe-a3b3-19f3d0b7d6a1",
      "name": "Embeddings Google Gemini",
      "type": "@n8n/n8n-nodes-langchain.embeddingsGoogleGemini",
      "position": [
        -608,
        320
      ],
      "parameters": {},
      "credentials": {
        "googlePalmApi": {
          "id": "3f4CCF4BMZnEfG6y",
          "name": "Google Gemini(PaLM) Api account"
        }
      },
      "typeVersion": 1
    },
    {
      "id": "05bbad6c-877c-4d6d-90e1-6c82d6560ae2",
      "name": "Vector Store simple1",
      "type": "@n8n/n8n-nodes-langchain.vectorStoreInMemory",
      "position": [
        -896,
        -544
      ],
      "parameters": {
        "mode": "insert",
        "memoryKey": {
          "__rl": true,
          "mode": "list",
          "value": "vector_store_key"
        }
      },
      "typeVersion": 1.3
    },
    {
      "id": "34948452-2e69-40cc-9b86-b78500873aab",
      "name": "Embeddings Google Gemini1",
      "type": "@n8n/n8n-nodes-langchain.embeddingsGoogleGemini",
      "position": [
        -896,
        -320
      ],
      "parameters": {},
      "credentials": {
        "googlePalmApi": {
          "id": "3f4CCF4BMZnEfG6y",
          "name": "Google Gemini(PaLM) Api account"
        }
      },
      "typeVersion": 1
    },
    {
      "id": "d6b2871c-78c6-4785-8913-262eb2364f7d",
      "name": "Chargeur de données par défaut",
      "type": "@n8n/n8n-nodes-langchain.documentDefaultDataLoader",
      "position": [
        -720,
        -400
      ],
      "parameters": {
        "options": {},
        "dataType": "binary",
        "textSplittingMode": "custom"
      },
      "typeVersion": 1.1
    },
    {
      "id": "6818e50a-ecc1-40e5-aac9-9d38fc85d3ec",
      "name": "Séparateur de texte récursif",
      "type": "@n8n/n8n-nodes-langchain.textSplitterRecursiveCharacterTextSplitter",
      "position": [
        -704,
        -256
      ],
      "parameters": {
        "options": {},
        "chunkOverlap": 200
      },
      "typeVersion": 1
    },
    {
      "id": "48da425a-c41f-4301-b4a7-df00f604ba5b",
      "name": "Requête HTTP",
      "type": "n8n-nodes-base.httpRequest",
      "position": [
        -1040,
        -448
      ],
      "parameters": {
        "url": "https://documents.iplt20.com/bcci/documents/1742707993986_Match_Playing_Conditions.pdf",
        "options": {}
      },
      "typeVersion": 4.2
    },
    {
      "id": "3fc9062b-fdef-421d-a7a3-d348c83cb51c",
      "name": "Au clic sur 'Exécuter le workflow'",
      "type": "n8n-nodes-base.manualTrigger",
      "position": [
        -1232,
        -448
      ],
      "parameters": {},
      "typeVersion": 1
    },
    {
      "id": "60491e32-d0c1-4e4a-922f-8ce976b481d1",
      "name": "Note adhésive",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -2576,
        -48
      ],
      "parameters": {
        "color": 6,
        "width": 2144,
        "height": 624,
        "content": "## Step 2\n## 2.1 Chat Trigger to initiate n8n native chat interface\n## 2.2 Simple Memory keeps the last 20 chat turns for context. This value can be edited within the node\n## 2.3 Simple Vector Store (retrieve-as-tool mode) receives the user’s query embedding, \n## finds the top-10 most relevant chunks stored in step 1, and supplies them as tool output. This will drive RAG\n**The name of vector store should match from Step 1, the embedding rule should match step 1\n## 2.4 Google Gemini Chat Model is the language model that is used as the llm model\n## 2.5 AI Agent orchestrates everything:\n** Uses the system prompt (“You are a cricket expert… If info is missing, say ‘Sorry I don’t know’”). to prompt the model\n** Has access to the memory (2.2) and the RAG tool (2.3).\n** Generates the final response with Google Gemini, strictly limited to the retrieved IPL cricket rules data.\n\n\n\n\n\n\n## Note: Google gemini API key credential needed\n##Using simple memory store nodes provided by n8n is the best way to get started to test out the workflow before you switch to more enterprise grade vector store nodes"
      },
      "typeVersion": 1
    },
    {
      "id": "1909411f-90b0-4cd5-823a-39f4f918cc5e",
      "name": "Note adhésive1",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -2576,
        -624
      ],
      "parameters": {
        "width": 2160,
        "height": 544,
        "content": "## Step 1\n## Load the reference material (run once via the Manual Trigger)\n## 1.1 Manual Trigger → HTTP Request downloads the IPL “Match Playing Conditions” PDF. \n## 1.2 Default Data Loader extracts text from the PDF.\n   **Type of data is binary\n## 1.3 Recursive Character Text Splitter breaks the text into overlapping chunks.\n   **This step ensures that the data chunks that are created in vector store have some overlap and hence less chance of hallucination\n   **Chunk size and chunk overlap are 2 variables to manage this \n## 1.4 Embeddings Google Gemini (1) converts each chunk to a vector.\n   **Connect the model with google gemini model. You will need your own api key for this\n   **Make note of the embedding model also since the same embedding model has to be selected in Step 2\n## 1.5 Simple Vector Store 1 inserts those vectors into an in-memory store under key\n   **Make note of the vector store name since it is same vector store you will have to use in Step 2\n\n\n## Note: Google gemini API key credential needed\n##Using Vector store nodes provided by n8n is the best way to get started to test out the workflow before you switch to more enterprise grade vector store nodes"
      },
      "typeVersion": 1
    },
    {
      "id": "63e38b73-3e30-47d7-86bb-afa2ad92dc2b",
      "name": "Note adhésive7",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -2576,
        -768
      ],
      "parameters": {
        "color": 5,
        "width": 2160,
        "height": 128,
        "content": "## This workflow has 2 Broad Steps\n## Step 1 - Vector store creation with set of ipl rules using Google Gemini Embedding. This will we used to drive RAG for model grouding    \n## Step 2 - Connecting the vector store with google gemini API model and enabling a chat interface to drive the chat bot\n"
      },
      "typeVersion": 1
    },
    {
      "id": "f45e2852-88a8-4f70-a124-01f2b06d9a19",
      "name": "Note adhésive2",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -1232,
        -544
      ],
      "parameters": {
        "color": 3,
        "width": 278,
        "height": 80,
        "content": "## Step 1.1"
      },
      "typeVersion": 1
    },
    {
      "id": "0b72e856-23c6-42c2-860e-8f761f861d95",
      "name": "Note adhésive3",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -608,
        -304
      ],
      "parameters": {
        "color": 3,
        "width": 166,
        "height": 128,
        "content": "## Step 1.2\n## Step 1.3"
      },
      "typeVersion": 1
    },
    {
      "id": "96c343b7-3961-49c1-97e0-35b4eee90d78",
      "name": "Note adhésive4",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -1088,
        -240
      ],
      "parameters": {
        "color": 3,
        "width": 150,
        "height": 80,
        "content": "## Step 1.4"
      },
      "typeVersion": 1
    },
    {
      "id": "f78516ba-4b17-4e48-9450-ba5d7cb123f1",
      "name": "Note adhésive5",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -592,
        -544
      ],
      "parameters": {
        "color": 3,
        "width": 150,
        "height": 80,
        "content": "## Step 1.5"
      },
      "typeVersion": 1
    },
    {
      "id": "b97281a4-6b1f-41a1-9a1e-c48be5a6854c",
      "name": "Note adhésive6",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -1248,
        96
      ],
      "parameters": {
        "color": 4,
        "width": 160,
        "height": 80,
        "content": "## Step 2.1"
      },
      "typeVersion": 1
    },
    {
      "id": "a8de0dce-eaa0-441d-b050-5374741f3b5f",
      "name": "Note adhésive8",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -976,
        464
      ],
      "parameters": {
        "color": 4,
        "width": 160,
        "height": 80,
        "content": "## Step 2.4"
      },
      "typeVersion": 1
    },
    {
      "id": "1f405862-c83e-4687-b919-3e128bcd2073",
      "name": "Note adhésive9",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -608,
        64
      ],
      "parameters": {
        "color": 4,
        "width": 160,
        "height": 80,
        "content": "## Step 2.3"
      },
      "typeVersion": 1
    },
    {
      "id": "dfb4cbe2-f6b0-45c4-bda7-d5f33a3b8e5f",
      "name": "Note adhésive10",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -800,
        464
      ],
      "parameters": {
        "color": 4,
        "width": 160,
        "height": 80,
        "content": "## Step 2.2"
      },
      "typeVersion": 1
    },
    {
      "id": "c5cfbb0b-2d09-40b8-ba18-5c4028d8a556",
      "name": "Note adhésive11",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        -928,
        -32
      ],
      "parameters": {
        "color": 4,
        "width": 160,
        "height": 80,
        "content": "## Step 2.5"
      },
      "typeVersion": 1
    }
  ],
  "active": false,
  "pinData": {},
  "settings": {
    "executionOrder": "v1"
  },
  "versionId": "98c130a5-eef0-4246-8a95-88a29c4e8ce6",
  "connections": {
    "48da425a-c41f-4301-b4a7-df00f604ba5b": {
      "main": [
        [
          {
            "node": "05bbad6c-877c-4d6d-90e1-6c82d6560ae2",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "15f7fbdc-ab77-4007-9a8e-8ddbe881d984": {
      "ai_memory": [
        [
          {
            "node": "352186bb-07d1-4d7d-9f0f-b57e0880fc11",
            "type": "ai_memory",
            "index": 0
          }
        ]
      ]
    },
    "d6b2871c-78c6-4785-8913-262eb2364f7d": {
      "ai_document": [
        [
          {
            "node": "05bbad6c-877c-4d6d-90e1-6c82d6560ae2",
            "type": "ai_document",
            "index": 0
          }
        ]
      ]
    },
    "dc61d50a-fdd8-4a21-974f-33aa8aab5c0a": {
      "ai_tool": [
        [
          {
            "node": "352186bb-07d1-4d7d-9f0f-b57e0880fc11",
            "type": "ai_tool",
            "index": 0
          }
        ]
      ]
    },
    "33d9a2a4-6f13-4cbe-a3b3-19f3d0b7d6a1": {
      "ai_embedding": [
        [
          {
            "node": "dc61d50a-fdd8-4a21-974f-33aa8aab5c0a",
            "type": "ai_embedding",
            "index": 0
          }
        ]
      ]
    },
    "69f8782c-c5d2-4693-bc00-a2ab58c61e08": {
      "ai_languageModel": [
        [
          {
            "node": "352186bb-07d1-4d7d-9f0f-b57e0880fc11",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "34948452-2e69-40cc-9b86-b78500873aab": {
      "ai_embedding": [
        [
          {
            "node": "05bbad6c-877c-4d6d-90e1-6c82d6560ae2",
            "type": "ai_embedding",
            "index": 0
          }
        ]
      ]
    },
    "4c32f558-efff-4eff-b714-202c7419a96c": {
      "main": [
        [
          {
            "node": "352186bb-07d1-4d7d-9f0f-b57e0880fc11",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "6818e50a-ecc1-40e5-aac9-9d38fc85d3ec": {
      "ai_textSplitter": [
        [
          {
            "node": "d6b2871c-78c6-4785-8913-262eb2364f7d",
            "type": "ai_textSplitter",
            "index": 0
          }
        ]
      ]
    },
    "3fc9062b-fdef-421d-a7a3-d348c83cb51c": {
      "main": [
        [
          {
            "node": "48da425a-c41f-4301-b4a7-df00f604ba5b",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  }
}

Foire aux questions

Comment utiliser ce workflow ?

Copiez le code de configuration JSON ci-dessus, créez un nouveau workflow dans votre instance n8n et sélectionnez "Importer depuis le JSON", collez la configuration et modifiez les paramètres d'authentification selon vos besoins.

Dans quelles scénarios ce workflow est-il adapté ?

Avancé - Ingénierie, IA Multimodale

Est-ce payant ?

Ce workflow est entièrement gratuit et peut être utilisé directement. Veuillez noter que les services tiers utilisés dans le workflow (comme l'API OpenAI) peuvent nécessiter un paiement de votre part.