🧠 Goal

Build a web app that interacts with OpenAI’s GPT-4 API using LangChain for prompt chaining and Streamlit for the UI, deployed on Azure.

🛠️ Tech Stack

• Azure OpenAI Service

• LangChain

• Python (FastAPI + Streamlit)

• Azure Web Apps (for deployment)

🧩 Steps

• Set Up Azure OpenAI

  1. Go to Azure Portal → Create resource → “Azure OpenAI”.

  2. Deploy a GPT-4 model and note your endpoint and key.

• Install LangChain & Streamlit

• pip install langchain streamlit openai

• .   Create the LangChain Script

• from langchain.llms import OpenAI
  from langchain.chains import LLMChain
  from langchain.prompts import PromptTemplate
  import os
  os.environ["OPENAI_API_KEY"] = "your-key"
  template = PromptTemplate.from_template("You are an AI assistant. Answer: {question}")
  chain = LLMChain(prompt=template, llm=OpenAI(temperature=0.7))
  def ask_question(q): return chain.run(question=q)

• Build Streamlit UI

• import streamlit as st
  st.title("Ask VishGPT")
  user_input = st.text_input("Your Question")
  if st.button("Ask"):
      response = ask_question(user_input)
      st.write(response)

• Deploy to Azure Web Apps

  • Use Azure CLI or GitHub Actions

  • Add .streamlit/secrets.toml for environment vars