Lecture Slides
Lecture decks for the chapters of Machine Learning Systems, to teach as written or to adapt for another course.
Introduction to Machine Learning Systems
The model on one machine: data, neural computation, architectures, frameworks, training, compression, hardware acceleration, benchmarking, serving, operations, and responsible engineering.
Volume IIScaling Machine Learning Systems
The fleet: compute infrastructure, networks, storage, collective communication, distributed training, fault tolerance, orchestration, serving at scale, security, sustainability, and responsible AI.
Volume IIIAgentic Machine Learning Systems
The trajectory: the foundation model call, test-time compute, context engineering, the KV cache, long-term memory, tool calling, sandboxes, security, the agent harness, durable execution, failure recovery, evaluation, operations, learning from trajectories, multi-agent coordination, and agent economics.
Volume IVPhysical AI: Machine Learning Systems That Sense and Act
The action that crosses into the physical world: the machine's body, brain, and nervous system, physical data, policy training, closed-loop evaluation, perception, spatial memory, grounded intent, planning, safety enforcement, workload placement, supervisory intervention, verification, release, and responsible physical AI.
TinyMLTiny Machine Learning
The HarvardX courseware: fundamentals, keyword spotting, visual wake words, anomaly detection, deployment on microcontrollers, and MLOps.
| The deck for projection, with builds that reveal a figure one step at a time. | |
| PowerPoint | The same slides for PowerPoint or Keynote. The notes pane holds the speaker notes, so presenter view shows what to say on each slide. |
| Speaker notes | What to say at each step, questions for the class with their expected answers, common misconceptions, and links to other lectures. |
| Source | The LaTeX Beamer source, the theme, and the scripts that draw every plot, so any slide can be changed and rebuilt. |
| License | CC BY-NC-SA 4.0. The decks may be taught, adapted, and shared for noncommercial use, with attribution and under the same license. |
Part of the MLSysBook Ecosystem
Textbook
A four-volume open textbook on machine learning systems.
Lecture Slides
Beamer lecture decks with speaker notes for each volume and TinyML.
Labs
Interactive Marimo notebooks that measure the book's claims.
Hardware Kits
Hands-on embedded ML deployment on real devices.
Instructor Hub
Course maps, syllabi, and adoption resources for teaching.
TinyTorch
A machine learning framework built from scratch, one module at a time.
MLSys·im
The analytical modeling engine behind the book's quantitative figures.