Build LLM App with Web Search: Step-by-Step Guide

Learn how to build your own powerful Large Language Model (LLM) application that can search the web, extract information, and answer your questions – all LOCALLY!

In this tutorial, I'll walk you through the process of creating a web-searching LLM from scratch, covering web crawling, vector databases, and semantic search. No API keys or cloud services needed!

---
🔥 Resources

Code: github.com/yankeexe/llm-app-with-web-search-demo
---
crawl4ai: docs.crawl4ai.com/
Duckduckgo_search: pypi.org/project/duckduckgo-search/
Streamlit: docs.streamlit.io/develop/api-reference
Ollama: github.com/ollama/ollama-python


---
⚡️ Follow me
Github: github.com/yankeexe
LinkedIn: www.linkedin.com/in/yankeemaharjan
Twitter (X): x.com/yankexe
Website: yankee.dev/

--
🎞️ Chapters

0:00 Intro
0:16 Application Flow
1:05 Application Demo
2:16 Project Setup
3:24 Build Frontend
5:32 Add Web Search
9:23 Checking robots.txt
11:52 Add Web Crawler
17:03 Setup Vector Database
28:37 Query Vector Database
31:17 Setup LLM
35:37 Application Demo
37:12 Out