Share Description
This course is an intermediate-level, project-based course, taught by Tom Chant, that places you on the bleeding edge of AI development. In just over 90 minutes, this course will guide you through building one of the most in-demand AI applications: a chatbot that can answer questions about a specific set of documents. You will learn to harness the power of the LangChain.js library and the LangChain Expression Language (LCEL) to create a "Retrieval-Augmented Generation" (RAG) system from scratch. The course project involves taking a source text, splitting it into manageable chunks, converting those chunks into vector embeddings using OpenAI, and storing them in a Supabase vector store. You will then build the chat logic to retrieve the most relevant information from your database and feed it to a large language model to generate an accurate, in-context answer. This is not a general-knowledge chatbot; it's a specialized AI assistant with deep knowledge of your provided data.
Topics the course covers
This course provides a dense, practical tour of the modern AI developer's toolkit. Key topics include:
- Data Processing: Using LangChain's TextSplitter to break down large documents into optimized chunks for processing.
- Embeddings & Vector Stores:
- Understanding the concept of vector embeddings.
- Using OpenAI's models to vectorize text chunks.
- Setting up and storing text vectors in a Supabase vector store.
- LangChain Expression Language (LCEL):
- Mastering the modern "pipe" (.pipe()) syntax to create chains.
- Building basic chains for simple data flow.
- Constructing complex, multi-step chains using RunnableSequence and RunnablePassthrough.
- Prompt Engineering:
- Creating dynamic prompt templates with input variables.
- Crafting specialized prompts for both standalone questions and answer generation.
- Retrieval & Generation:
- Implementing a retriever to fetch relevant documents from the vector store based on a user's query.
- Using StringOutputParser to get a clean, usable response from the language model.
- Building the App:
- Wiring up the complete chatbot logic.
- Integrating chat memory to allow for contextual, follow-up questions.
- Troubleshooting common performance issues.
Benefits of opting for this course
By completing this course, you will:
- Build a complete, portfolio-ready RAG chatbot—one of the most common and valuable real-world AI applications.
- Learn the official, modern LangChain.js library and its powerful Expression Language (LCEL), putting you ahead of developers using older methods.
- Gain hands-on experience with the full AI development lifecycle: data processing, vector storage, retrieval, and generation.
- Master the integration of multiple key technologies: LangChain.js, OpenAI, and Supabase.
- Develop a "true superpower" for developers: the ability to give AI deep, contextual knowledge about any private data.
- Solidify your skills through Scrimba's interactive challenges, building the "muscle memory" needed to become a proficient AI engineer.
Why take this course
If you're a web developer with a solid grasp of JavaScript and you're ready to move into the world of AI, this is the perfect next step. This course is not just a theoretical overview; it's a hands-on-keyboard guide to building a scalable, real-world application. The skills you'll learn here are at the heart of the generative AI revolution and are what companies are actively hiring for. You will learn to build applications that go beyond simple API calls to ChatGPT, creating systems that can reason over private documentation, power internal help desks, or analyze specific knowledge bases. Taught by Tom Chant and created in partnership with LangChain, this official course provides a direct, efficient path to mastering a critical AI skill set, transforming you from a developer who uses AI into one who builds with it.
