Coursera Bets $100 Million That Andrew Ng Can Do What Coursera Can’t

Class Central is learner-supported. When you buy through links on our site, we may earn an affiliate commission.

Coursera

Build a Local AI Assistant with LLMs

JetBrains via Coursera

Go to class Write review

Overview

Google, IBM & Meta Certificates – 40% Off
One plan covers every Professional Certificate on Coursera.
Unlock All Certificates
Build your own AI assistant that answers questions from your documents – entirely on your local machine. Assuming a basic acquaintance with Python, this course will teach you how to run a local LLM, turn PDFs into searchable chunks, generate embeddings, store them in a vector database, and connect retrieval and generation into a complete RAG (Retrieval-Augmented Generation) pipeline. You’ll create OpenAI-compatible and RAG endpoints with FastAPI, work with Ollama and Qdrant, and finish by building a browser-based interface for asking questions and reviewing sources. This course stands out because everything is built locally, from end to end. Instead of relying on black-box cloud services, you will master every step of the system you build – from document processing and vector search to prompt construction and answer generation. You’ll learn by actually building, adding one piece at a time. With each module, you’ll unlock a new feature in your project., By the end, you will have a production-ready AI project you can run, customize, and share.

Syllabus

  • Week 1
    • In this module, you’ll set up your local environment and make your first call to a large language model. You’ll learn what LLMs are, why local inference is crucial for privacy and experimentation, and how Docker helps you run the system reliably across different machines. You’ll explore the core architecture of the course project, understand the API layer and the role of HTTP and JSON, and write your first Python function that talks to the model. By the end of this module, you’ll have a working local LLM request flowing through your own API.
  • Week 2
    • In this module, you’ll build and configure the API layer that powers the project. You’ll learn how FastAPI structures endpoints, request models, and validation, how the server communicates with Ollama, and why the OpenAI chat format has become the industry standard. You’ll also work with async requests, system prompts, multi-turn conversations, and automated testing using pytest and FastAPI’s TestClient. By the end of this module, you’ll understand how the chat API works from request to response and how to verify it with tests.
  • Week 3
    • In this module, you’ll move from model calls to document processing. You’ll extract text from PDFs, see why PDF parsing is harder than it looks, and learn how to split long documents into retrieval-friendly chunks. You’ll compare chunk sizes, master overlap strategies, and build a document pipeline that attaches critical metadata like source file names and chunk positions. By the end of this module, you’ll be able to turn any PDF into structured chunks that are ready for indexing.
  • Week 4
    • In this module, you’ll make text searchable by meaning. You’ll learn what embeddings are, how cosine similarity measures semantic closeness, and why vector databases differ from traditional databases. Then, you’ll connect these ideas in code by generating vectors with an embedding model, storing them in Qdrant, and implementing the indexing endpoint that ties document chunks to vector storage. By the end of this module, your documents will be embedded, stored, and ready for retrieval.
  • Week 5
    • In this module, you’ll close the RAG loop. You’ll explore the two core phases of RAG – retrieval and generation – and see how a user’s question becomes a vector, how relevant chunks are selected, and how prompt structure guides the model to answer strictly from context. Then, you’ll implement the query endpoint, tune parameters like Top-K and score thresholds, and return answers with clear sources. By the end of this module, your system will answer questions accurately, grounded entirely in your indexed documents.
  • Week 6
    • In this module, you’ll give your project a browser interface and make it demo-ready. You’ll learn how to add a lightweight web UI with HTMX and FastAPI, render templates and HTML fragments, and connect the form-based interface to the same RAG logic you built earlier. You’ll also see how to test the full user flow in the browser and turn your backend project into something easy to share with others. By the end of this module, you’ll have a complete local AI assistant with a usable web interface.

Taught by

JetBrains Academy team

Tags

Reviews

5.0 rating, based on 8 Class Central reviews

4.9 rating at Coursera based on 12 ratings

Start your review of Build a Local AI Assistant with LLMs

  • Anonymous
    This course is an excellent, practical introduction to running LLMs locally. JetBrains did a great job breaking down the setup process step by step, from configuring the environment to building a working assistant. The hands-on approach made it easy to follow along even without prior experience in local AI deployment. I especially liked how it covered real-world considerations like performance and resource management. Highly recommended for developers wanting to get started with local AI tools.
  • Anonymous
    It was a fun course by Jetbrains, and I learned a lot from it including how to run llms locally to help you in your daily tasks or as you want to use those llms along with RAG which helps and restricts llm to only respond from provided documents, so they never hallucinate.
  • Profile image for Fabio Rossetti
    Fabio Rossetti
    Good hands on course on how to use local llms via docker containers to implement a simple RAG system in python. The assignments are also easily doable with the aid of coding agents, if the need arises.
  • Anonymous
    Its help me to understand how llm works, and the video explanation pretty good

    I hope the women instructor voices louder, sometimes I need to manually adjust my volume because of it
  • Anonymous
    It was surely great course 😀, enjoyed building the agent and best part build it locally was the fun part and felt like a weapon in pocket 😁
  • Anonymous
    Very good course for learning how do build a local AI assisant and implemeting RAG, it really demystify the concept of local ai.
  • Anonymous
    Great course. Very practical and engaging. Loved that in the end you have a ready project in your hands. Thanks JetBrains!
  • Anonymous
    it was nice experience, the assignment submission thingy was cool...
    could have been more in depth / harder
Go to class

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Sign up for free
Someone learning on their laptop while sitting on the floor.