RAGForge: Build Your Own Local AI Knowledge Assistant

Ever wondered how AI can answer questions from your own documents? What if you could run an AI model locally and build an application that can understand and retrieve information from your own data?

RAGForge is a hands-on workshop organized by FOSS-SRM that takes you from the fundamentals of Large Language Models (LLMs) to building your own Retrieval-Augmented Generation (RAG) application.

In this workshop, you will explore how modern AI applications work under the hood and learn how different components come together to create a document-aware AI assistant.

What you’ll explore:

• Understanding Large Language Models (LLMs)

• Running LLMs locally using Ollama

• Working with Qwen

• Processing and chunking documents

• Understanding and generating embeddings

• Storing and retrieving information using vector databases

• Understanding semantic search and retrieval

• Building a complete RAG pipeline

• Integrating the LLM and RAG pipeline into an application

• Creating a user-friendly interface using Streamlit

What you’ll build

By the end of the workshop, you’ll build a local AI knowledge assistant that can take a document, retrieve relevant information from it, and use an LLM to generate meaningful answers to your questions.

The complete pipeline will look like:

Document → Chunking → Embeddings → Vector Database → Retrieval → Qwen → Answer

No prior experience with RAG is required. Basic familiarity with Python and a willingness to learn and experiment is recommended.

This is a hands-on workshop, so bring your laptop, be ready to code, experiment, break things, fix them, and most importantly — build something of your own.

Come curious. Leave with your own AI application.

Organized by FOSS-SRM.