Building an Neural Network From Scratch - Understanding the Gears Behind AI

Modern AI can seem mysterious, but at its core, it’s built on simple mathematical principles repeated at scale. In this talk, we’ll build a small neural network from scratch — no TensorFlow, no PyTorch — just TypeScript and pixels. Along the way, we’ll unpack how weights, activation functions, and back-propagation work together to form the foundation of systems capable of recognizing faces and powering today’s AI models.

5:30 - 6:00 - Socializing/Networking
6:00 - Speaker begins