Valerii Konchin

Renzu Select is an application built for professional photographers who need to process thousands of RAW files after a shoot. I designed, developed, and shipped this product to the App Store entirely on my own as a solo developer. The app loads embedded JPEG previews from RAW files (.NEF, .RAF, .CR2/CR3, .ARW, .DNG) instantly, so photographers can start culling without waiting for heavy editors to launch.

I implemented an on-device ML pipeline for smart stacking. The app analyzes photos using vector embeddings and visual-similarity clustering to group similar shots automatically. All AI processing runs locally on the Mac with zero cloud dependency. Other features I built include a 5-star rating system with XMP metadata support, dynamic viewing themes, a non-blocking import pipeline with background caching, automatic session recovery, and a professional export system with naming templates and metadata presets. All heavy processing runs in isolated background threads to keep the UI responsive.

Neru Plan is a meal planning app for iOS and iPad that I designed, developed, and shipped to the App Store as a solo developer. The name comes from the Japanese verb "neru", which means "to cultivate a plan carefully." The app uses Apple intelligence for on-device AI recipe imports: users can snap a photo of a recipe or paste a link, and the app extracts and structures all ingredients and steps locally. No data ever leaves the device, and the app works entirely offline.

I built a custom ingredient matching system, which I call "Atoms", according to which ingredients aren't stored as plain text. "Tomato" in a soup and "Tomatoes" in a salad automatically merge into one item on the shopping list. The shopping list engine handles automatic aggregation and unit conversion between cups, tablespoons, and metric measurements. Other features include an infinite-scroll calendar with drag-and-drop meal planning, dynamic serving size scaling that recalculates shopping lists without modifying original recipes, and iCloud sync between iPhone and iPad. I also built a recipe import parser supporting 10+ popular cooking websites. All AI processing uses Apple Foundation Models running locally.

Moku View is a native macOS storage analyzer built with Swift and SwiftUI. It visualizes disk space usage through a hierarchical tree and a synchronized pie chart. It is a lightweight, performance-focused alternative to TreeSize, designed for identifying large files and folders.