TinyML From Scratch

TinyML From Scratch is a practical, hands-on introduction to Tiny Machine Learning and Edge AI on resource-constrained embedded devices. Designed for beginners, students, engineers, and makers, this book guides readers step-by-step through the complete TinyML workflow, from sensor data acquisition to model training and real-time inference on microcontrollers.

Using the Arduino Nano 33 BLE Sense and Edge Impulse, readers build real TinyML applications involving motion recognition, sound classification, sensor fusion, anomaly detection, and embedded AI systems. The book emphasizes practical implementation, production-oriented thinking, and understanding the engineering trade-offs between accuracy, latency, memory usage, and energy efficiency.

Unlike theory-heavy machine learning books, this book focuses on learning by building. Every chapter includes real projects, deployment workflows, and hands-on experimentation with real sensor data.

Whether you are starting your TinyML journey or exploring embedded AI for practical applications, TinyML From Scratch provides a clear and accessible path into the rapidly growing world of intelligent low-power devices and Edge AI systems.