☁️ Lesson 9.4: Deployment Options (Local, Web, Cloud Hosting)

 

☁️ Lesson 9.4: Deployment Options (Local, Web, Cloud Hosting)

Module 9: Tools, Plugins, and Deployment
Course: Build Your Own AI Chat System (Like ChatGPT or DeepSeek) Using Free Open-Source Tools

🎯 Lesson Objective:

By the end of this lesson, learners will be able to deploy their chatbot in different environments—locally, on the web, or in the cloud—depending on their needs and technical ability.

🧠 Why Deployment Matters

Building a chatbot is great—but deployment is what makes it usable by:

  • Clients

  • Team members

  • Students or the public

Whether you’re showcasing a portfolio, building internal tools, or selling a product, deployment gives your chatbot a real home.

🚀 Deployment Options Overview

Method Description Best For
Localhost Runs on your own computer Development & testing
Internal Server / Intranet Local network deployment Small teams & private orgs
Free Cloud Hosting Limited cloud usage Portfolios, MVPs
Full Cloud Deployment Scalable, global access Businesses & production apps

1️⃣ Local Deployment (Your Own Computer)

Tools:

  • Streamlit, Gradio, Ollama, LangChain

Steps:

streamlit run app.py
# OR
python chatbot.py

Pros:

  • Free

  • Great for testing

Cons:

  • Only works while your PC is on

  • No public access unless you use tunneling

Bonus Tip: Use ngrok to expose locally:

ngrok http 8501

2️⃣ Team-Only Deployment (Internal Server / LAN)

Host on a Raspberry Pi, Ubuntu server, or virtual machine on your network.

Steps:

  • Install required tools (Python, Ollama, etc.)

  • Run chatbot on a fixed port

  • Access using local IP (e.g., 192.168.1.100:8501)

Pros:

  • No internet dependency

  • Good for sensitive company data

Cons:

  • Only accessible on same network unless VPN is set up

3️⃣ Free Cloud Hosting Options

Platform Supports Limitations
Streamlit Cloud Python/Streamlit apps Limited compute, 1GB
Hugging Face Spaces Gradio + Streamlit 3 GB RAM, slow LLMs
Render Web services Sleep after inactivity

Example: Deploy to Streamlit Cloud

  • Push your app to GitHub

  • Go to streamlit.io/cloud

  • Connect GitHub repo

  • Add requirements.txt (e.g., streamlit, langchain, ollama, chromadb)

  • Deploy!

4️⃣ Full Cloud Deployment (Production-Ready)

Use platforms like:

  • AWS EC2 + FastAPI

  • DigitalOcean + Docker

  • Railway, Vercel, Heroku (paid plans)

Why Use This:

  • Custom domains (e.g., chat.yoursite.com)

  • Team collaboration

  • Can scale with user growth

Common Setup:

  1. Dockerize your chatbot

  2. Add API authentication

  3. Use background tasks (e.g., Celery) for async tools

  4. Add SSL, monitoring, and autoscaling

🧪 Try It Yourself Activity

Task: Choose one deployment method and put your chatbot online or on a local network.

Optional Challenges:

  • Use ngrok to share your localhost bot temporarily

  • Push to GitHub and deploy via Streamlit Cloud

  • Set up a Raspberry Pi and deploy your AI assistant as a private server for home or office

💡 Tips & Best Practices

  • For quick demos, use Streamlit Cloud or Hugging Face

  • Use .env files for API keys and secrets (never commit to GitHub!)

  • Keep requirements.txt clean and lightweight

  • Monitor memory usage, especially if using local LLMs

  • Back up your vector store (Chroma, FAISS) regularly

🧠 Recap:

You’ve now learned how to:

  • Deploy your chatbot locally for personal use

  • Host it privately for a team or internal project

  • Share it publicly using free cloud services

  • Prepare it for scaling on production cloud infrastructure

  •  

85
PreviousNext