LogicBridge Podcast
On-Device AI for iOS
Sandboxed is a podcast for iOS developers who want to add AI and machine learning features to their apps—without needing a PhD in ML. Each episode, we take one practical topic—like Vision, Core ML, or Apple Intelligence—and walk through how it works, what you can build with it, and how to ship it.
Latest Episodes
Your First ML-Powered App
The 'Hello World' of AI isn't a print statement. We walk through the engineering reality of integrating the Vision Framework, Core ML, and the Zero-Copy pipeline.
When to Use ML (And When Not To)
The mark of a Senior Engineer is knowing when to reject AI. We explore the 'Scalpel vs Sledgehammer' decision matrix, the hardware tax of inference, and why Regex beats BERT for parsing.
Privacy: Apple's AI Differentiator
The Physics of Private AI. We deconstruct the App Sandbox, Secure Enclave, and the thermal constraints that force efficient, private model design.
Apple's ML Stack Overview
Core ML, Create ML, Vision, NL — the 'Onion' of Apple frameworks. Learn when to use the Power Tools (Vision) and when to build a Custom Engine (Core ML).
The Compiler & The Chip: How Code Becomes Thinking
We strip away the 'Interpreted AI' myth to reveal the reality of the Core ML Compiler. We explore Layer Fusion, the ANE's Systolic Array architecture, the 'Ping-Pong' performance trap, and why shipping a .mlpackage is actually shipping a blueprint for a silicon factory.
Neural Networks: The 10,000-Foot View
Demystifying deep learning by replacing biological metaphors with structural ones. We explore the Corporate Hierarchy, ReLU, and why Inference is just a Forward Pass.
Training vs. Inference
Learn the critical difference between teaching a model (Training) and using it (Inference), and why on-device AI relies on the latter.
What Is a Model?
Understand the .mlmodel format, weights and biases, and why models are frozen artifacts—not code you can debug.
Types of ML Tasks
Master Core ML tasks—classification, regression, detection, and segmentation—and match the task to the hardware to avoid thermal and UX traps.
How Machines "Learn" from Data
How weights, biases, and loss actually work—and how to keep Core ML models from overfitting or learning shortcuts.
The ML Landscape: Cloud vs. On-Device
Why Apple bets on on-device AI: The strategic trade-off between infinite cloud compute and zero-latency privacy.
What Is Machine Learning, Really?
Demystifying ML for iOS developers—what it can and can't do in your apps.