How to Build an AI Chatbot Using OpenAI and Streamlit

How to Build an AI Chatbot Using OpenAI and Streamlit

Want to build a production-level AI Chatbot Application that goes beyond a simple chat interface? This full-stack Python project uses FastAPI as the backend, Streamlit as the frontend, OpenAI GPT-3.5 Turbo as the AI brain, and MongoDB as the database — making it one of the most complete final year projects for BCA, MCA, and B.Tech CS/IT students in 2026.

This project includes 7 powerful AI modules — from AI chat and image generation to PDF Q&A and text summarization — all secured with JWT authentication and daily usage limits.

Project Overview

Project Name	AI Chatbot Application
Backend	FastAPI (Python)
Frontend	Streamlit
AI Engine	OpenAI GPT-3.5 Turbo + DALL-E 3
Database	MongoDB with Motor async driver
Authentication	JWT Token + bcrypt
Difficulty	Intermediate to Advanced
Best For	BCA, MCA, B.Tech CS/IT Final Year Students

7 Key Modules of This Project

Module	Description
AI Chat	Conversational AI using GPT-3.5 Turbo with full context memory
Image Generator	Generate images from text prompts using DALL-E 3
PDF Q&A	Upload any PDF and ask questions — get AI-powered answers
Text Summarizer	Paste any text and get short, medium, or long summary instantly
Chat History	Browse, view, resume, or delete all past conversations
Profile and Settings	Edit username, change password, monitor daily usage stats
Authentication	Secure login and registration with JWT tokens and bcrypt hashing

Download Full Source Code

The complete source code — backend routes, Streamlit pages, .env template, and requirements.txt — is available below.

Technologies Used

Layer	Technology	Purpose
Frontend	Streamlit	Interactive web UI
Backend	FastAPI	High-performance async REST API
AI Chat	OpenAI GPT-3.5 Turbo	Chat, PDF Q&A, and text summarization
Image AI	OpenAI DALL-E 3	Text-to-image generation
Database	MongoDB	Store users, conversations, usage data
DB Driver	Motor	Async MongoDB driver for Python
Auth	python-jose + bcrypt	JWT tokens and secure password hashing
PDF Parser	PyPDF2	Extract and chunk text from uploaded PDFs
HTTP Client	Requests	Frontend-to-backend communication
Config	python-dotenv	Load environment variables from .env

Project Folder Structure

Ai chatbot/
├── .env
├── requirements.txt
├── run.bat
│
├── backend/
│   ├── main.py
│   ├── config.py
│   ├── database/
│   │   └── mongodb.py
│   ├── models/
│   │   ├── user.py
│   │   └── chat.py
│   ├── routes/
│   │   ├── auth.py
│   │   ├── chat.py
│   │   ├── image.py
│   │   ├── pdf.py
│   │   └── summarizer.py
│   └── services/
│       ├── auth_service.py
│       ├── ai_service.py
│       └── pdf_service.py
│
└── frontend/
    ├── app.py
    ├── pages/
    │   ├── login.py
    │   ├── register.py
    │   ├── chat.py
    │   ├── image_gen.py
    │   ├── pdf_qa.py
    │   ├── summarizer.py
    │   ├── history.py
    │   └── settings.py
    └── utils/
        └── api_client.py

Prerequisites

• Python 3.10 or higher installed on your system
• MongoDB running on localhost:27017 (local install or Docker)
• OpenAI API Key from platform.openai.com with billing enabled
• Basic knowledge of Python, REST APIs, and terminal commands

How to Run This Project

Step 1 — Install All Dependencies

cd "Ai chatbot"
pip install -r requirements.txt

Step 2 — Configure Your .env File

Create a .env file in the root project folder and add the following:

OPENAI_API_KEY=sk-proj-your-actual-api-key-here
MONGODB_URL=mongodb://localhost:27017
DATABASE_NAME=ai_chatbot
JWT_SECRET_KEY=your-random-secret-key-here
JWT_ALGORITHM=HS256
ACCESS_TOKEN_EXPIRE_MINUTES=60
MAX_MESSAGES_PER_DAY=50
MAX_MESSAGE_LENGTH=2000
MAX_PDF_SIZE_MB=10

Step 3 — Start MongoDB

# Local install
mongod

# Or using Docker
docker run -d -p 27017:27017 --name mongodb mongo:latest

Step 4 — Start the FastAPI Backend

cd "Ai chatbot"
uvicorn backend.main:app --host 127.0.0.1 --port 8000

Backend runs at: http://localhost:8000
API docs at: http://localhost:8000/docs

Step 5 — Start the Streamlit Frontend

streamlit run frontend/app.py --server.port 8501

Frontend runs at: http://localhost:8501 — open this in your browser and you’re ready!

API Endpoints Reference

Authentication

Method	Endpoint	Description	Auth
POST	/api/auth/register	Register a new user	No
POST	/api/auth/login	Login and get JWT token	No
GET	/api/auth/profile	Get user profile	Yes
PUT	/api/auth/profile	Update username or name	Yes
POST	/api/auth/change-password	Change password	Yes

AI Chat

Method	Endpoint	Description	Auth
POST	/api/chat/send	Send message, get AI response	Yes
GET	/api/chat/conversations	List all conversations	Yes
GET	/api/chat/conversations/{id}	Get full conversation	Yes
DELETE	/api/chat/conversations/{id}	Delete a conversation	Yes
GET	/api/chat/usage	Get today’s usage stats	Yes

Image, PDF and Summarizer

Method	Endpoint	Description	Auth
POST	/api/image/generate	Generate image using DALL-E 3	Yes
POST	/api/pdf/upload	Upload a PDF file	Yes
POST	/api/pdf/ask	Ask question about the PDF	Yes
POST	/api/summarize/	Summarize a block of text	Yes

Security Features

• JWT Authentication — all protected routes require a valid Bearer token
• bcrypt Password Hashing — passwords are never stored as plain text
• Pydantic Input Validation — all request data is validated for format and length
• Daily Rate Limiting — each user is limited to 50 messages per day (configurable)
• PDF File Size Cap — uploads are limited to 10MB by default
• CORS Middleware — cross-origin requests handled securely

Configurable Limits via .env

Limitation	Default	Config Key
Messages per day per user	50	MAX_MESSAGES_PER_DAY
Max characters per message	2,000	MAX_MESSAGE_LENGTH
Max PDF file size	10 MB	MAX_PDF_SIZE_MB
Max text for summarization	10,000 chars	hardcoded in model
Image prompt max length	1,000 chars	hardcoded in model
JWT token expiry	60 mins	ACCESS_TOKEN_EXPIRE_MINUTES

MongoDB Collections

Collection	Purpose	Key Fields
users	User accounts and profiles	username, email, password_hash, full_name
chats	Conversations and messages	user_id, title, module, messages[], pdf_text
daily_usage	Rate limit tracking	user_id, date, count

How the Application Works

Login and Registration

1. User fills the form on the Streamlit UI
2. Frontend sends POST to /api/auth/register or /api/auth/login
3. Backend validates credentials and creates or verifies user in MongoDB
4. JWT token is returned and stored in Streamlit session state
5. All future requests include the token as Authorization: Bearer token

AI Chat

1. User types a message in the Streamlit chat UI
2. Message is sent to /api/chat/send with JWT token
3. Backend loads the last 10 messages for conversation context
4. Full context is sent to OpenAI GPT-3.5 Turbo
5. AI response is saved to MongoDB and returned to the frontend

PDF Q&A

1. User uploads a PDF — PyPDF2 extracts all text from the file
2. Extracted text is stored in a MongoDB conversation document
3. User asks a question about the PDF content
4. Text is chunked and relevant sections found using keyword matching
5. Relevant context plus question is sent to GPT-3.5 Turbo and answer is returned

Image Generation

1. User enters a text prompt and selects size and quality settings
2. Request is sent to /api/image/generate
3. Backend calls DALL-E 3 API with the prompt and preferences
4. Generated image URL is returned and displayed in the Streamlit UI

Why This is a Great Final Year Project

• Uses real industry-level tools — FastAPI, OpenAI API, MongoDB, JWT
• Covers both frontend and backend — ideal for full stack Python developers
• 7 modules in one project — highly impressive for viva and presentations
• Easy to extend — add voice input, language translation, or analytics
• GPT-3.5 Turbo and DALL-E 3 are the most in-demand AI skills in 2026
• One-click Windows startup with run.bat — easy to demo to examiners

Need help setting it up? Drop a comment below or visit updategadh.com for more free Python projects with source code.

---