Deep-dives into building production AI agents, offline-first mobile apps, zero-alloc Go backends, and multi-theme design systems.
A deep dive into low-latency voice transcription with Whisper, screen perception with YOLOv8, and dual-engine reasoning between local Ollama models and the Gemini 2.0 Flash API.
How we built deterministic streak calculation logic, optimistic local updates, and conflict-free cloud replication in HabitFlow without latency penalties.
Adapting foundation models to understand informal economies, Mobile Money (MoMo) SMS payloads, and communal savings structures across East Africa.
Design engineering principles for building interfaces that feel tactile, intentional, and comfortable during prolonged night sessions.
Why SQLite paired with Go's standard library net/http creates lightning-fast, ultra-low-memory backend microservices that cost $0 to run in production.
Reach out directly to discuss architecture, open-source builds, or engineering collaborations.