Waiwai Mengmeng publishes a book to teach you how to create "local LLM applications". Behind Mengbao's shy appearance is actually a master's degree in information science.

The new idol of young people, Wai Wai Meng Meng, Wai Bao releases a book to teach you how to play with AI.

Waiwai Mengmeng is a well-known internet celebrity and singer-songwriter in Taiwan. The single "Wai Meng's 300 Days Gift" has over 140,000 views on YouTube. The main creator is Mengbao Mengbert, a graduate of the Department of Civil Engineering at National Taiwan University, and is currently studying in the graduate program of the Department of Music at Fu Jen Catholic University. The one publishing this book is Weibao Weibert, who holds a master's degree in Computer Science from National Taipei University. He has won the gold medal in the Ministry of Education's National University Artificial Intelligence Competition AI CUP and the Trend Micro Student Pre-employment Award.

This time he published a book under the pen name "好崴寶(Weibert Weiberson)", combining theory and practice, releasing the book titled "Local End Ollama × LangChain × LangGraph × LangSmith Development Manual: Building RAG, Agent, SQL Applications."

Building local large language model applications through practical learning

WaiBao states that this book is suitable for developers who want to build large language models locally, teaching you how to create applications of large language models using local devices. The book includes practical cases, allowing you to directly use the code inside for implementation. It also uses the daily life of WaiWai and MengMeng for practical applications, such as using everyday language for semantic similarity retrieval in vector databases.

The book emphasizes "data does not flow out" and "models change as you like." By running large language models on the local end through the Ollama system, it not only avoids the upload of sensitive data to the cloud but also saves on costly API fees. Combined with the LangChain application framework, developers can quickly transform existing functions into AI tools, and even switch between models without having to rewrite code, significantly enhancing flexibility.

In addition, the book also fully introduces the application of RAG ( for retrieval-augmented generation ), allowing quick import of data from PDFs, web pages, and JSON files to build a dedicated knowledge base for the model. Furthermore, with the process control of AI Agent and LangGraph, developers can design intelligent agents that can autonomously plan actions while avoiding infinite loops. Finally, through the LangSmith evaluation tool, the design of prompts, performance metrics, and task decisions during the development process can be clearly quantified and tracked.

( 2025 Comprehensive Analysis of the Top 5 Mainstream LLMs: Understanding Payment, Applications, and Security All at Once )

This article by Wai Wai and Meng Meng teaches you how to create "local LLM applications". Behind Meng Bao's shy appearance is actually a master's degree in computer science, first appearing in Chain News ABMedia.