In today\u2019s digital world, businesses and individuals aim to provide instant and accurate answers to website visitors. With increased demand for seamless communication, AI-driven chatbots<\/a> have become a crucial tool for user interaction and offering useful information in a split second. Chatbots can search, comprehend, and utilize website data efficiently, making customers satisfied and enhancing the customer experience for companies. In this article, we will explain how to build a chatbot that fetches information from a website, processes it efficiently, and engages in meaningful conversations with the assistance of Qwen-2.5<\/a>, LangChain<\/a>, and FAISS. we will learn the main components and integration process.<\/p>\n This article was published as a part of the\u00a0<\/strong><\/em>Data Science Blogathon.<\/strong><\/em><\/a><\/p>\n Many businesses struggle with handling large volumes of customer queries efficiently. Traditional customer support teams often face delays, leading to frustrated users and increased operational costs. Moreover, hiring and training support agents can be expensive, making it difficult for companies to scale effectively.<\/p>\n A chatbot helps by offering instant and automated responses to user questions without needing a human. Businesses are able to cut support costs considerably, increase customer interaction, and provide users with instant answers to their questions. AI-based chatbots are capable of handling large volumes of data, determining the right information in a matter of seconds, and reacting correctly based on the context, making them very beneficial for businesses nowadays.<\/p>\n Website chatbots are mostly used in E-learning platforms, E-commerce websites, customer support<\/a> platforms, and news websites<\/a>.<\/p>\n Also Read: Building a Writing Assistant with LangChain and Qwen-2.5-32B<\/a><\/em><\/p>\n Here\u2019s a flowchart explaining the working of our chatbot.<\/p>\n Now let\u2019s see how to build a custom website chatbot using Qwen-2.5-32b, LangChain, and FAISS.<\/p>\n Let\u2019s begin by setting up the prerequisites.<\/p>\n Paste the API key in .env file.<\/p>\n Now let\u2019s get into the actual coding part.<\/p>\n Ensures compatibility with Windows by setting the correct event loop policy for asyncio, as Windows uses a different default event loop.<\/p>\n This disables Streamlit\u2019s file watcher to improve performance by reducing unnecessary file system monitoring.<\/p>\n Stores processed text chunks in FAISS . Then Converts the FAISS vectorstore into a retriever that can fetch relevant chunks based on user queries.<\/p>\n This step creates a qa_chain <\/b>by linking the retriever (FAISS) with the LLM, ensuring responses are based on retrieved website content.<\/p>\n This displays previous conversation messages.<\/p>\n This code handles user input, retrieval, and response generation in a\u00a0 chatbot. When a user enters a message, it is stored in st.session_state.conversation<\/b> and displayed in the chat interface. A loading spinner appears while the chatbot processes the query using qa_chain.invoke({\u201cinput\u201d: user_input})<\/b>, which retrieves relevant information and generates a response. The assistant\u2019s reply is extracted from the response dictionary, ensuring a fallback message if no answer is found. Finally, the assistant\u2019s response is stored and displayed, maintaining a smooth and interactive chat experience.<\/p>\n Get the full code on GitHub here<\/a><\/p>\n Now let\u2019s try out a few prompts on the chatbot we just built.<\/p>\n Prompt: <\/strong>\u201cCan you list some ways to engage with the Analytics Vidhya community?\u201d<\/em><\/p>\n Response<\/strong>:<\/p>\n Prompt: <\/strong>\u201cwhat other programs they offers?\u201d<\/em><\/p>\n Response<\/strong>:<\/p>\n AI chatbots have transformed the manner in which people communicate on the internet. Using advanced models like Qwen-2.5-32b, businesses and individuals can make sure that their chatbot responds well and suitably. As technology continues to advance, using AI chatbots on websites will be the order of the day, and people will be able to access information easily.<\/p>\n In the future, developments like having long conversations, voice questioning, and interacting with bigger knowledge pools can further advance chatbots even more.<\/p>\n The media shown in this article is not owned by Analytics Vidhya and is used at the Author\u2019s discretion.<\/strong><\/p>\n A. It uses UnstructuredURLLoader from LangChain to extract content from specified URLs.<\/p>\n<\/p><\/div>\n A. FAISS (Facebook AI Similarity Search) helps store and retrieve relevant text chunks efficiently based on user queries.<\/p>\n<\/p><\/div>\n A. The chatbot uses Groq\u2019s Qwen-2.5-32B, a powerful LLM, to generate answers based on retrieved website content.<\/p>\n<\/p><\/div>\n A. Yes! Simply modify the urls list to include more websites, and the chatbot will fetch, process, and retrieve information from them.<\/p>\n<\/p><\/div>\n A. It follows a Retrieval-Augmented Generation (RAG) approach, meaning it retrieves relevant website data first and then generates an answer using LLM.<\/p>\n<\/p><\/div>\n<\/p><\/div>\nLearning Objectives<\/h4>\n
\n
Why Use a Website Chatbot?<\/h2>\n
Key Components of the Chatbot<\/h4>\n
\n
How Does This Chatbot Work?<\/h2>\n
![\"How](\"https:\/\/cdn.analyticsvidhya.com\/wp-content\/uploads\/2025\/03\/chatbot_flow.webp\")
<\/figure>\n<\/div>\nBuilding a Custom Chatbot Using Qwen-2.5-32b and LangChain<\/h2>\n
Step 1: Setting Up the Foundation<\/h3>\n
1. Environment Setup<\/h4>\n
# Create a Environment\npython -m venv env\n\n# Activate it on Windows\n.\\env\\Scripts\\activate\n\n# Activate in MacOS\/Linux\nsource env\/bin\/activate<\/code><\/pre>\n2. Install the Requirements.txt<\/h4>\n
pip install -r https:\/\/raw.githubusercontent.com\/Gouravlohar\/Chatbot\/refs\/heads\/main\/requirements.txt<\/code><\/pre>\n3. API Key Setup<\/h4>\n
![\"API](\"https:\/\/cdn.analyticsvidhya.com\/wp-content\/uploads\/2025\/03\/image_aQNQvbZ.webp\")
<\/figure>\nAPI_KEY=\"Your API KEY PASTE HERE\"<\/code><\/pre>\nStep 2: Handling Windows Event Loop (For Compatibility)<\/h3>\n
import sys\nimport asyncio\n\nif sys.platform.startswith(\"win\"):\n asyncio.set_event_loop_policy(asyncio.WindowsSelectorEventLoopPolicy())<\/code><\/pre>\nStep 3: Importing Required Libraries<\/h3>\n
import streamlit as st\nimport os\nfrom dotenv import load_dotenv<\/code><\/pre>\n\n
os.environ[\"STREAMLIT_SERVER_FILEWATCHER_TYPE\"] = \"none\" <\/code><\/pre>\nStep 4: Importing LangChain Modules<\/h3>\n
from langchain_huggingface import HuggingFaceEmbeddings \nfrom langchain_community.vectorstores import FAISS\nfrom langchain_community.document_loaders import UnstructuredURLLoader\nfrom langchain.text_splitter import RecursiveCharacterTextSplitter\nfrom langchain_groq import ChatGroq\nfrom langchain.chains import create_retrieval_chain\nfrom langchain.chains.combine_documents import create_stuff_documents_chain\nfrom langchain_core.prompts import ChatPromptTemplate<\/code><\/pre>\n\n
Step 5: Loading Environment Variables<\/h3>\n
load_dotenv()\ngroq_api_key = os.getenv(\"API_KEY\")\n\nif not groq_api_key:\n st.error(\"Groq API Key not found in .env file\")\n st.stop()<\/code><\/pre>\n\n
\u00a01. Function to Load Website Data<\/h4>\n
def load_website_data(urls):\n loader = UnstructuredURLLoader(urls=urls)\n return loader.load()<\/code><\/pre>\n\n
2. Function to Chunk Documents<\/h4>\n
def chunk_documents(docs):\n text_splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=50)\n return text_splitter.split_documents(docs)<\/code><\/pre>\n\n
3. Function to Build FAISS Vector Store<\/h4>\n
def build_vectorstore(text_chunks):\n embeddings = HuggingFaceEmbeddings(model_name=\"sentence-transformers\/all-MiniLM-L6-v2\")\n return FAISS.from_documents(text_chunks, embeddings)<\/code><\/pre>\n\n
4. Function to Load\u00a0Qwen-2.5-32b<\/h4>\n
def load_llm(api_key):\n return ChatGroq(groq_api_key=api_key, model_name=\"qwen-2.5-32b\", streaming=True)<\/code><\/pre>\n\n
5. Streamlit UI Setup<\/h4>\n
st.title(\"Custom Website Chatbot(Analytics Vidhya)\")<\/code><\/pre>\n6. Conversation History Setup <\/h4>\n
if \"conversation\" not in st.session_state:\n st.session_state.conversation = []<\/code><\/pre>\n\n
Step 6: Fetching and Processing Website Data<\/h3>\n
urls = [\"https:\/\/www.analyticsvidhya.com\/\"]\ndocs = load_website_data(urls)\ntext_chunks = chunk_documents(docs)<\/code><\/pre>\n\n
Step 7: Building FAISS Vector Store<\/h3>\n
vectorstore = build_vectorstore(text_chunks)\nretriever = vectorstore.as_retriever()<\/code><\/pre>\nStep 8: Loading the Groq LLM<\/h3>\n
llm = load_llm(groq_api_key)<\/code><\/pre>\nStep 9:\u00a0Setting Up Retrieval Chain<\/h3>\n
system_prompt = (\n \"Use the given context to answer the question. \"\n \"If you don't know the answer, say you don't know. \"\n \"Use detailed sentences maximum and keep the answer accurate. \"\n \"Context: {context}\"\n)\n\nprompt = ChatPromptTemplate.from_messages([\n (\"system\", system_prompt),\n (\"human\", \"{input}\"),\n])\n\ncombine_docs_chain = create_stuff_documents_chain(llm, prompt)\nqa_chain = create_retrieval_chain(\n retriever=retriever,\n combine_docs_chain=combine_docs_chain\n)<\/code><\/pre>\n\n
Step 10: Displaying Chat History<\/h3>\n
for msg in st.session_state.conversation:\n if msg[\"role\"] == \"user\":\n st.chat_message(\"user\").write(msg[\"message\"])\n else:\n st.chat_message(\"assistant\").write(msg[\"message\"])<\/code><\/pre>\nStep 11: Accepting User Input<\/h3>\n
user_input = st.chat_input(\"Type your message here\") if hasattr(st, \"chat_input\") else st.text_input(\"Your message:\")<\/code><\/pre>\n\n
Step 12: Processing User Queries <\/h3>\n
if user_input:\n st.session_state.conversation.append({\"role\": \"user\", \"message\": user_input})\n if hasattr(st, \"chat_message\"):\n st.chat_message(\"user\").write(user_input)\n else:\n st.markdown(f\"**User:** {user_input}\")\n\n with st.spinner(\"Processing...\"):\n response = qa_chain.invoke({\"input\": user_input}) \n\n \n assistant_response = response.get(\"answer\", \"I'm not sure, please try again.\")\n st.session_state.conversation.append({\"role\": \"assistant\", \"message\": assistant_response})\n\n if hasattr(st, \"chat_message\"):\n st.chat_message(\"assistant\").write(assistant_response)\n else:\n st.markdown(f\"**Assistant:** {assistant_response}\")<\/code><\/pre>\nFinal Output<\/h3>\n
![\"Custom](\"https:\/\/cdn.analyticsvidhya.com\/wp-content\/uploads\/2025\/03\/image_ic06IxR.webp\")
<\/figure>\nTesting the Chatbot<\/h2>\n
![\"Custom](\"https:\/\/cdn.analyticsvidhya.com\/wp-content\/uploads\/2025\/03\/image_yql1cfm.webp\")
<\/figure>\n![\"output](\"https:\/\/cdn.analyticsvidhya.com\/wp-content\/uploads\/2025\/03\/image_6ERB1n0.webp\")
<\/figure>\nConclusion<\/h2>\n
Key Takeaways<\/h4>\n
\n
Frequently Asked Questions<\/h2>\n
![\"Gourav](\"https:\/\/av-eks-lekhak.s3.amazonaws.com\/media\/lekhak-profile-images\/converted_image_b56toW7.webp\")
<\/p>\n