Machine learning might seem intimidating, but it doesn’t have to be.
This hands-on training takes you from ““What is machine learning?”” to running Neural Networks and working with pre-trained models, all within the Elixir ecosystem. No advanced math degree required, just curiosity and good Elixir skills.
We’ll start with the fundamentals: Linear Regression that actually makes sense, then progressively build up to NN (Neural Networks), Computer Vision, and NLP (Natural Language Processing). Each concept comes with live coding examples you can run on your laptop, turning abstract mathematical ideas into concrete, working code.
You’ll discover how Nx brings tensor operations to the BEAM, how Axon makes neural network training surprisingly elegant, and how Bumblebee lets you work with large language models and run speech-to-text with Whisper, all with just a few lines of code. While we’ll use Elixir, Nx, and Axon, the core concepts and skills you learn will transfer seamlessly to other ML frameworks like PyTorch, Keras, and TensorFlow.
By the end, you’ll have trained models to recognize handwritten digits, classify images, predict sequences, and run powerful pre-trained transformers for text and speech.
This isn’t just theory. It’s about giving you the confidence and practical knowledge to bring machine learning into your Elixir projects. Come ready to experiment, learn, and maybe be a little amazed at what we can teach silicon to do.
Duration
- 8 hours
Tutorial objectives: By the end of this training, the participants will:
- Understand machine learning fundamentals without getting overwhelmed by complex mathematics.
- Work with the Elixir ML ecosystem including Nx, Axon, and Bumblebee for building and running models.
- Build working models from linear regression to neural networks, with hands-on coding throughout.
- Integrate pre-trained models like LLMs and Whisper into Elixir applications.
- Evaluate and improve model performance using appropriate metrics and techniques.
- Walk away with practical code examples ready to adapt for real-world ML projects, plus transferable skills that apply to Python frameworks like PyTorch, Keras, and TensorFlow
Target audience:
- Elixir developers who are curious to understand how machine learning actually works, who want to add AI capabilities to their applications but feel intimidated by the mathematical complexity typically associated with ML. Ideal for backend developers expanding into AI/ML, Elixir enthusiasts exploring the BEAM ML ecosystem, and anyone interested in practical, hands-on learning that demystifies machine learning concepts.
Skills required:
- Intermediate Elixir knowledge
- Basic math (no advanced statistics required)
Software prerequisites:
- Elixir >=1.17, Erlang/OTP 27+
- Livebook installed and working