RAG and Chatbots with LangChain

Comfortable Python and the ideas from the Generative AI course: tokens, prompts and embeddings. If you have done those two courses you are ready.

No. Embeddings, the re-ranker and FAISS run locally, and hosted models are available on OpenRouter's free tier, which has daily request limits. A long scored run (about 55 calls for thirty questions with a judge) can pass those limits, so use the fully local build with Ollama or spread the run over several days. Ollama needs nothing online once the models are downloaded.

There are two projects. The first is a manual-reading assistant over three vendor manuals. It has citations, a re-ranker, a thirty-question evaluation and a Streamlit page. The second is an assistant over a year of maintenance work orders. It routes text questions to RAG and counting questions to templated SQL.

Because a RAG demo is easy and a RAG tool is not. In the first project the service desk sets its own limits, such as recall above 0.85 and refusal at 1.0. Those are that scenario's numbers, not a standard. The course teaches you to set limits from the users and to prove you reach them.

Yes. Every module, practice task, project and assessment. You create an account so your progress is saved and the assessments can be marked. The certificate is the only paid item, and only if you want it.

An EDWartens Certificate of Completion, issued when you have finished the modules and passed the final assessment (the project is optional practice), with a verification code anyone can check. It is not a vendor credential and is never described as one.

The creators named in the Credits block at the foot of this page, on their own YouTube channels. EDWartens did not make the videos and the creators are not affiliated with EDWartens. What EDWartens wrote is the study plan, the notes, the practice tasks, the projects and the assessments.

No. The lab is optional practice. You can skip it whenever you like, and it never blocks your progress, the final assessment or your certificate.

Yes. Every lab level for this course opens in your browser, free, with nothing to install.

The RAG and Chatbots with LangChain course has 12 modules. It opens with "What RAG is, and what it fixes", goes through "LangChain: models, prompts, chains", "The basic RAG chain and chat memory", "Re-ranking and advanced indexing" and "Build, serve, keep private" and ends with "Final assessment". The full list with minutes is under "What you learn" on this page.

After the RAG and Chatbots with LangChain course: Load, clean and chunk vendor manuals so parameter tables survive; Embed locally, search with hybrid BM25 + vector retrieval and re-rank with a cross-encoder; Build a RAG chain that cites (doc p.N), refuses with NOT FOUND and handles follow-ups. These are the practical outcomes the course is written around, and the final assessment of 15 questions checks them.

For the RAG and Chatbots with LangChain course, software: Google Colab or a laptop; LangChain, sentence-transformers, FAISS; a free OpenRouter key or Ollama. Hardware: none. A laptop with 8 GB RAM for the fully local build. The course itself is free.

RAG and Chatbots with LangChain is rated intermediate. Intermediate. Needs Python (the Python for AI course) and the Generative AI and LLM Foundations course, or equivalent. It is written for engineers, developers and students applying AI to real work, and takes about 15.1 hours at your own pace.

Skills from the RAG and Chatbots with LangChain course (for example: load, clean and chunk vendor manuals so parameter tables survive) are used in roles such as AI engineer, machine learning engineer, data analyst and automation developer, and in machine learning engineer work. The course gives you practical evidence of the skill; it does not promise a job, and employers decide on experience and interview.