Build local LLM applications using Python and Ollama

Name: Build local LLM applications using Python and Ollama
Rating: 4.6 (988 reviews)

Learn to create LLM applications in your system using Ollama and LangChain in Python | Completely private and secure

Role Play

Created byStart-Tech Academy, Pukhraj Parikh, Abhishek Bansal

Last updated 4/2026

English

German [Auto],English [Auto],

What you'll learn

Download and install Ollama for running LLM models on your local machine
Set up and configure the Llama LLM model for local use
Customize LLM models using command-line options to meet specific application needs
Save and deploy modified versions of LLM models in your local environment
Develop Python-based applications that interact with Ollama models securely
Call and integrate models via Ollama’s REST API for seamless interaction with external systems
Explore OpenAI compatibility within Ollama to extend the functionality of your models
Build a Retrieval-Augmented Generation (RAG) system to process and query large documents efficiently
Create fully functional LLM applications using LangChain, Ollama, and tools like agents and retrieval systems to answer user queries

Course content

6 sections • 27 lectures • 2h 0m total length

Introduction6:12
Learn to build local llm applications with Python and Ollama, install and run models on your own system for privacy, explore LangChain, retrieval augmented generation, and agentic tools.
Downloading and Installing Ollama4:35
install Ollama on your system, download a local model, and interact via the GUI or command line; then explore Python libraries to build local LLM applications.
Setting up Ollama and downloading Llama LLM model8:05
Learn to set up Ollama, download a llama model via the command line, and balance model size against performance and local resource use for fast on-premises prompts.
This is a milestone3:52
Celebrate reaching this milestone in building local llm apps. You are among the top 50% of learners; stay motivated with resources, q&a, ai assistant help, subtitles, and your certificate.
Model Customization options in CMD or terminal7:11
Explore command line customization with Ollama: enter a model, use slash commands to view options, and configure verbose and system messages to control responses, including token metrics.
Creating, saving and using a modified Ollama model3:40
Learn to create, save, and load a modified ollama model named Jarvis AI, view model details with /show, and manage parameters, system messages, and session history including /clear.
Quiz

Installing and Setting up Python4:52
Install and set up Python and the development environment, choose an IDE, and use Jupyter Notebook to write and run code in cells with pip install notebook and ctrl+enter.
Using Ollama library in Python7:59
Install and import the ulama library in Python, pull and manage models with ulama, and generate or chat with models using Python, including explanations of system, user, and assistant roles.
Calling the Model using Ollama Rest API3:23
Call the Ollama rest api to query a locally downloaded model by creating a client with the host, selecting a model, and sending messages to receive a response.
Ollama OpenAI Compatibility5:02
Learn how Ollama enables OpenAI compatibility by using the same code to call local llama models via a base URL, enabling OpenAI-style requests, Python workflows, and LangChain integration.
Quiz

What is LangChain and why are we using it5:04
LangChain sits between user prompts and LLMs, enabling easy model replacement and features like memory, knowledge integration, and agents, including retrieval augmented generation and tool use.
Basics of Langchain - Prompt Templates and LLM Models5:35
Basics of Langchain - Formatting the output4:41
Learn to build a Lang chain by creating a prompt template, invoking a language model, and formatting the output with a string output parser to extract content.
Quiz

Concept of Retrieval Augmented Generation RAG System3:00
What is the RAG Process5:43
Master the retrieval augmented generation process: prepare knowledge as text chunks, create contextual embeddings, store them in a vector database, and retrieve relevant data to power lm-based answers.
Loading and Chunking the document using LangChain and Ollama6:24
Load the document with a text loader, split it into chunks with a recursive character splitter, embed with Ulama via LangChain, and store in Chroma vector database for query-based retrieval.
Embedding chunks using LangChain and Ollama4:46
Convert text chunks into embeddings using local models (llama embeddings, nomic embed text), store them in a chroma vector database, and use similarity search with LangChain and Ollama.
Building complete RAG application for answering user questions - Part 14:31
Build a rag application that answers user questions using retrieved context. Create a prompt from the context and question, send it to a llama 3.1 model, and parse the output.
Building complete RAG application for answering user questions - Part 24:20
Build a RAG application by retrieving relevant documents, creating a context with format docs, and feeding it to a prompt for the model; Lang Chain provides embeddings across providers.
Quiz
About the upcoming role play.0:38
Explaining the RAG Process and Applications in a Job Interview

Understanding Tools and Agents5:38
Explore tools and agents with llama models, learning to create tools and agents in Python, and to call travel portals using predefined LangChain tools for dynamic querying.
Tools calling with LangChain and Llama3.15:27
Explore tool calling with LangChain and lemma 3.1 using inbuilt tools like DuckDuckGo search and Wikipedia to identify models that support tools and build a tool calling agent.
Agents using LangChain and Llama3.15:40
Define a LangChain agent with an LM model, tools, and prompt, then run an agent executor to call tools and obtain local LM results via DuckDuckGo and Wikipedia.
Quiz
The final milestone!1:33
Build local LLM applications using Python and Ollama by reaching the final milestone, join the top 5% of students, and download your certificate to your registered email.
About the upcoming Roleplay0:42
Interview Practice

Requirements

Basic Python knowledge is recommended, but no prior AI experience is required.

Description

If you are a developer, data scientist, or AI enthusiast who wants to build and run large language models (LLMs) locally on your system, this course is for you. Do you want to harness the power of LLMs without sending your data to the cloud? Are you looking for secure, private solutions that leverage powerful tools like Python, Ollama, and LangChain? This course will show you how to build secure and fully functional LLM applications right on your own machine.

In this course, you will:

Set up Ollama and download the Llama LLM model for local use.
Customize models and save modified versions using command-line tools.
Develop Python-based LLM applications with Ollama for total control over your models.
Use Ollama's Rest API to integrate models into your applications.
Leverage LangChain to build Retrieval-Augmented Generation (RAG) systems for efficient document processing.
Create end-to-end LLM applications that answer user questions with precision using the power of LangChain and Ollama.

Why build local LLM applications? For one, local applications ensure complete data privacy—your data never leaves your system. Additionally, the flexibility and customization of running models locally means you are in total control, without the need for cloud dependencies.

Throughout the course, you’ll build, customize, and deploy models using Python, and implement key features like prompt engineering, retrieval techniques, and model integration—all within the comfort of your local setup.

What sets this course apart is its focus on privacy, control, and hands-on experience using cutting-edge tools like Ollama and LangChain. By the end, you’ll have a fully functioning LLM application and the skills to build secure AI systems on your own.

Ready to build your own private LLM applications? Enroll now and get started!

Who this course is for:

Software developers who want to build and run private LLM applications on their local machines.
Data scientists looking to integrate advanced LLM models into their workflow without relying on cloud solutions.
Privacy-focused professionals who need to maintain complete control over their data while leveraging powerful AI models.
Tech enthusiasts interested in exploring local LLM setups using cutting-edge tools like Ollama and LangChain.

Build local LLM applications using Python and Ollama

What you'll learn

Explore related topics

Course content

Getting started with local models6 lectures • 34min

Using Ollama with Python4 lectures • 21min

Using LangChain in Python for LLM applications3 lectures • 15min

Building Retrieval Augmented Generation - RAG applications8 lectures • 29min

Building Tools and Agents based applications6 lectures • 19min

Conclusion2 lectures • 2min

Requirements

Description

Who this course is for: