A good RAG course takes you from “I called an embedding API” to “I can debug why retrieval returned the wrong chunk.” These five free DeepLearning.AI short courses cover the full retrieval stack: vector storage, embeddings-to-application pipelines, building and evaluating RAG systems, agent evaluation, and the accuracy work that turns a demo green.
Organized by stack layer: storage, then application, then retrieval-plus-evaluation, then accuracy. Five courses, each a distinct layer, no overlap.
Vector databases
Building Vector Databases with Pinecone (Pinecone, via DeepLearning.AI, free)
The storage layer. You learn what a vector index actually is (HNSW, flat, IVF), how Pinecone organizes namespaces and metadata filters, and the trade-offs between managed vector databases and rolling your own with FAISS. The right course when you are about to choose a vector store and want to understand what you are buying. Starts at the data layer, not the chatbot.
Course: Building Applications with Vector Databases.
Embeddings to applications
Vector Databases from Embeddings to Apps (Pinecone, via DeepLearning.AI, free)
The application layer on top of the storage course. This one picks up where the first leaves off: how to generate embeddings at scale, how to chunk documents so retrieval actually works, and how to wire a semantic-search front end to a vector store. If the first course is “what is a vector database,” this one is “how do I build a product on one.” Take both if you are shipping retrieval for the first time.
Course: Vector Databases from Embeddings to Applications.
Building and evaluating RAG
Building and Evaluating Advanced RAG Applications (DeepLearning.AI, free)
The most complete single course on this list. Covers the full RAG pipeline (chunking, embedding, retrieval, generation) and then spends real time on evaluation: how to measure whether retrieved context is relevant, whether the answer is grounded, and whether the system is actually better than no retrieval at all. The evaluation half is what most tutorials skip and what separates a working RAG system from a hallucination machine.
Course: Building and Evaluating Advanced RAG Applications.
Evaluating AI Agents (DeepLearning.AI, free)
Listed here for its retrieval-evaluation angle: the same eval discipline (LLM-as-judge, trajectory evaluation, regression suites) applies to agentic RAG pipelines where the agent decides what to retrieve. If you are building a tool-using agent that calls a retriever, this course covers how to score it. (Also featured in our agent engineering roundup for its primary agent-eval framing.)
Course: Evaluating AI Agents.
Improving accuracy
Improving Accuracy of LLM Applications (DeepLearning.AI, free)
The accuracy layer. Once retrieval works, accuracy is the next bottleneck: prompt engineering, model selection, output parsing, and the feedback loops that catch regressions before users do. This course is short (under two hours) and tactical. The right pick when your RAG demo works on five examples and fails on the sixth, and you need a debugging method, not another API.
Course: Improving Accuracy of LLM Applications.
If you only do one: Building and Evaluating Advanced RAG Applications. It covers the most ground and is the only course here that treats evaluation as a first-class topic, which is where most RAG projects actually fail.