Building Your Second Brain AI Assistant Using Agents, LLMs and RAG

This is an open-source course, part of Decoding AI's open-source series, that teaches how to build production-ready generative-AI systems using large language models, retrieval-augmented generation, agents and LLMOps. Its six modules build an end-to-end AI assistant that chats with a Second Brain, a personal knowledge base of notes, resources and stored material, based on a concept from Tiago Forte.

By the end of the course, a learner can architect and implement a production-ready agentic RAG and LLM system from scratch: designing production-ready LLM architectures, applying LLMOps and software-engineering best practices, fine-tuning and deploying LLMs, and using industry tools such as OpenAI, Hugging Face, MongoDB, ZenML, Opik and Unsloth. The course uses Notion as its example knowledge source and provides a curated AI and machine-learning resource list, but its pipeline is adaptable to other sources such as Google Drive or Calendar, and no Notion account is required to follow along.

It is written as Jupyter notebooks and released under the MIT license, built in collaboration with MongoDB, Comet, Opik, Unsloth and ZenML.