Ever wondered how Buddhist practitioners know which days their prayers carry the most spiritual weight?
We built the Karma Kagyu Calendar to answer precisely that: a lifestyle app that brings ancient lunar wisdom into your pocket.
The opportunity Our client needed a beautiful, functional mobile app for both iOS and Android. As an AI-First company, this greenfield project was the perfect chance to prove our AI-First methodology, introduced just this September. We wanted to examine not just how AI can answer questions, but how much of the actual implementation it could handle.
How we built it We leaned heavily into AI-assisted development with GitHub Copilot and Claude Code. The real game-changer? Connecting Figma directly to our LLM through the Figma MCP integration. The AI agent could read our designs, implement components, and suggest solutions, all of which were validated and refined by our developers before integration. Using React with Capacitor, we built once and deployed everywhere, no separate iOS and Android codebases needed.
What the app does Users see daily multipliers based on the Buddhist lunar calendar—those special days when prayers and good deeds carry more spiritual significance (we're talking multipliers like 100 million). The app shows exactly when the lunar day begins so practitioners can plan accordingly. Daily notifications keep routines on track, and detailed astrological insights guide each day.
What we learned The Karma Kagyu Calendar proved something important: AI tools aren't just helpful for answering development questions; they can complete concrete development steps that, after thorough review, integrate seamlessly into the codebase. The notification system, for example, was AI-generated mainly, then validated and optimized by our team.
The key? Developer expertise remains essential. AI accelerates and augments, but the final judgment call is always human. That's the real win: our team now has more time for thoughtful planning and refining solutions, instead of getting bogged down in repetitive implementation tasks. Have you experimented with AI-assisted development on a real project yet?