FastAPI integration patterns in LangChain - Mini Project: Build & Apply

FastAPI Integration Patterns with LangChain

📖 Scenario: You are building a simple web API using FastAPI that integrates with LangChain to process text inputs and return generated responses. This project will guide you through setting up the data, configuration, core logic, and final API endpoint.

🎯 Goal: Create a FastAPI app that accepts a text input, uses LangChain to generate a response, and returns the result as JSON.

📋 What You'll Learn

Create a FastAPI app instance

Set up a LangChain prompt template

Implement an endpoint that accepts POST requests with text input

Use LangChain to generate a response based on the input text

Return the generated response in JSON format

💡 Why This Matters

🌍 Real World

This project shows how to build a simple web API that uses LangChain to process natural language inputs and return generated responses, useful for chatbots, question answering, or AI assistants.

💼 Career

Understanding FastAPI integration with LangChain is valuable for backend developers working on AI-powered web services, enabling them to build scalable and maintainable APIs.

1

DATA SETUP: Create FastAPI app and import LangChain

Import FastAPI from fastapi and PromptTemplate from langchain.prompts. Then create a FastAPI app instance called app.

LangChain

from fastapi import FastAPI
from langchain.prompts import PromptTemplate

# Create FastAPI app instance
# Your code here

Need a hint?

Use app = FastAPI() to create the app instance.

2

CONFIGURATION: Define a prompt template for LangChain

Create a PromptTemplate instance called template with the template string 'Answer the question: {question}' and input variable list ['question'].

LangChain

from fastapi import FastAPI
from langchain.prompts import PromptTemplate

app = FastAPI()

# Define LangChain prompt template
# Your code here

Need a hint?

Use PromptTemplate(template='Answer the question: {question}', input_variables=['question']).

3

CORE LOGIC: Create a POST endpoint to process input text

Define a POST endpoint function called generate_answer with path '/generate' that accepts a JSON body with a question string. Use the template.format(question=question) to generate the response string called answer.

LangChain

from fastapi import FastAPI
from langchain.prompts import PromptTemplate
from fastapi import Request

app = FastAPI()

template = PromptTemplate(template='Answer the question: {question}', input_variables=['question'])

# Create POST endpoint to generate answer
# Your code here

Need a hint?

Use @app.post('/generate') decorator and async function with request.json().

4

COMPLETION: Add CORS middleware for cross-origin requests

Import CORSMiddleware from fastapi.middleware.cors. Add CORS middleware to app allowing origins ['*'], methods ['*'], and headers ['*'].

LangChain

from fastapi import FastAPI
from langchain.prompts import PromptTemplate
from fastapi import Request
from fastapi.middleware.cors import CORSMiddleware

app = FastAPI()

# Add CORS middleware
# Your code here

template = PromptTemplate(template='Answer the question: {question}', input_variables=['question'])

@app.post('/generate')
async def generate_answer(request: Request):
    data = await request.json()
    question = data['question']
    answer = template.format(question=question)
    return {'answer': answer}

Need a hint?

Use app.add_middleware(CORSMiddleware, allow_origins=['*'], allow_methods=['*'], allow_headers=['*']).