AI

Practical AI for app makers: on-device models, LLM features, and AI-assisted development.

Adding Speech-to-Text to a Small App: On-Device or Cloud?

Voice input looks like a single API call, but the real choices are on-device or cloud, streaming or batch, and how you handle permissions, silence and bad transcripts. Here is a practical way to choose and ship it in a small app.

AI Coding Assistants in a Solo Workflow: Guardrails That Work

An AI coding assistant can speed up a solo developer, but it can also leave you with code you don't understand. Here are practical guardrails: which tasks to hand off, how to give the assistant context, how to review its diffs, and how to tell whether it's actually helping.

Designing LLM Features That Fail Gracefully in a Small App

Model calls time out, hit rate limits, return broken JSON or confidently wrong answers. Here is a practical way to plan for those failures in a small app: a fallback for each one, timeouts based on measured latency, output checks, clear UI states and a remote kill switch.

Adding an On-Device Model to Your Mobile App: A Practical Checklist

On-device models can make AI features private, offline-capable and free per request. They also bring app size, device coverage and quality trade-offs. Here is a practical checklist for deciding whether to run a model locally, and how to ship it without breaking older phones.

Episode

00:00:00 00:00:00