R&D

Qintu is an end-to-end Kotlin Multiplatform experiment built to ensure unified data transfer objects (DTOs) and shared structures across the entire distributed stack. In complex multi-language systems, coordinating data shapes between separate backend and client languages often introduces major schema bottlenecks; Kotlin solves this by sharing identical data models directly from the Ktor backend right through to the frontend clients.

Beyond the technical challenge, the project stems from our team's shared interest in ancient Chinese metaphysical systems — specifically the mathematical frameworks of the I Ching, Ba Gua, and Wu Xing. Translating these traditional structures into pure, deterministic code was a process we simply enjoyed.

Structurally, it features a classic yarrow-stalk probability calculator, local encryption using SQLCipher for user privacy, and is built to expose a structured Chinese metaphysics API, enabling personal AI agents to query readings directly and bypass traditional graphical interface constraints.

QDarte is a travel discovery and listing project for Latin America, designed to surface emerging places, local businesses, and regional content. It began with a practical question: can a lean platform make useful destination information easier to find without the overhead of a large editorial operation?

The research focus is a lightweight automated operating model. Routine sourcing, initial enrichment, editorial preparation, and verification workflows are designed to run through focused pipelines, with human oversight where it is useful.

Under the hood, a Signal Engine monitors social and map trend momentum, a local AI pipeline compiles candidate listings, and a magic-link Verification Engine supports operator confirmation through WhatsApp.