Linux Training+32 (0)2 747 47 01Registration FormMail

Embedded Linux Penguin image Practical Artificial Intelligence on Linux

Course Overview
This course provides a hands-on introduction to Artificial Intelligence with a strong focus on Linux-based development environments and OpenSource tooling. Rather than treating AI as a black box, we explore how modern AI systems actually work. From classical neural networks to OpenSource Large Language Models (LLMs) and how they can be practically implemented, deployed and operated on Linux systems in a large environment. Participants implement, deploy and operate OpenSource Large Language Models (LLMs) on Linux, with an emphasis on production use, data locality and system integration. Classical neural networks are covered using Python, TensorFlow and Keras, enabling participants to combine LLM-based systems with traditional machine learning workflows. Throughout the course, the emphasis is on understanding, transparency, safeguards and reproducibility, using OpenSource AI components on Linux rather than closed or cloud-locked platforms. During this course all participants will have the opportunity to build and experiment with a multifunctional small-footprint embedded target with hardware-optimised AI capabilities. After the course, participants can take the board with them to continue experimenting.

Knowledge prerequisites
IT background and general Linux skills. Linux Power User or similar command line experience. Basic Python knowledge is required as well. Some programming experience is advised to better understand the concepts.

Method
Course / Workshop, classical education combined with practical exercises.

Participants
Engineers, developers and technical professionals responsible for designing, building or evaluating Artificial Intelligence systems on Linux.

Course Flow

1. Introduction
  • Foundations of Artificial Intelligence on Linux
  • Overview of modern AI system architectures
  • Linux-based AI development environments
  • Python fundamentals (as needed)

2. OpenSource Large Language Models (LLMs)

  • LLM architectures and capabilities
  • Selecting and evaluating OpenSource LLMs
  • Local deployment on Linux (CPU / GPU)
  • Production-oriented setups and configuration
  • Data locality, privacy and compliance considerations
  • Security boundaries and trust models for LLMs

3. LLM Integration, Safeguards and Operation

  • Prompt engineering for deterministic and controlled behaviour
  • Input validation and output filtering
  • Safeguards against prompt injection and data leakage
  • Retrieval-Augmented Generation (RAG) with access control
  • Tool calling and function integration using least-privilege principles
  • LLM-based agents: capabilities, limits and containment
  • Secure integration with Linux services and APIs

4. Classical Neural Networks

  • Multilayer Perceptron (MLP)
  • Convolutional Neural Networks (CNN)
  • Recurrent Neural Networks (RNN)
5. Neural Networks with Python, TensorFlow and Keras
  • Model definition and training
  • Evaluation and debugging
  • Transfer learning and pre-trained models
  • Combining neural networks with LLM-based systems


6. Applied Machine Learning and Data Workflows

  • Data preparation and pipelines
  • Prediction and estimation tasks
  • Hybrid ML + LLM architectures


7. Limitations, Debugging and Operational Considerations

  • Performance and resource constraints
  • Failure modes and limitations of LLMs
  • Monitoring, updates and maintenance
  • Discussion of participant use cases


Administrative Information
Course Dates:
On Demand

Courseware:
Course materials provided, complemented with one book: Tutorial: Practical Artificial Intelligence on Linux, by Jasper Nuyens. A free ARM-based Embedded Linux board with camera and an Intel NCSM2450.DK1 Movidius Neural Compute Stick are included.

Price 3.150,- € + VAT

More information
Phone: +32 (0)2 747 47 01
Course Registration Form

Back to Linux Belgium Education

You can find the Full Calendar here.

Copyright 2001-2023, Linux Belgium. All rights reserved. Linux is a trademark of Linus Torvalds.
Linux Belgium is a registered trademark of Linux Belgium b.v.b.a. All other trademarks are the property of their respective owners.