Sharing the journey of building the best dictation app for privacy and productivity. Slow grind, until it all moves fast. Learning, mistakes and all.
You have a fixed amount of attention each day. You're spending it before you even start working. On mental bandwidth, the cost of constant input, and why the shower is the last place your brain gets to think.
The things nobody tells you about organic growth, word of mouth, and why your parents still don't understand what you do. Reflections from a solo founder on the patterns Paul Graham describes that hit uncomfortably close to home.
The gap between wanting to build something and actually building it is not motivation. It's something smaller and more embarrassing. On productive procrastination, the Kaizen method, and why two minutes of real work beats three weeks of planning.
Two speech engines, one app, and the optimization that halved startup time. A practical guide to managing CoreML model loading, memory budgets, and the tradeoffs between streaming and chunk-based transcription.
How to build invisible application awareness in under 100ms. A deep dive into detecting email contexts on macOS using bundle IDs, window title matching, Accessibility APIs, and Vision OCR—all running on-device.
How a one-line counter increment hid a distributed systems bug. A debugging story about Cloudflare Workers, KV Store eventual consistency, and why idempotent event handlers aren't optional—even at small scale.
Yakki.ai