VIA
FeaturedApp de formación multilingüe con compañeros de IA
Una plataforma de formación católica en 4 idiomas con 5 compañeros de IA (Anthropic API), liturgia diaria, canto gregoriano con detección de tono en vivo y donaciones vía Stripe. El producto técnicamente más complejo de mi portafolio — multiagente, ML de audio, pagos, internacionalización completa — diseñado, construido y lanzado en solitario.
Problema
Traditional faith formation relies on in-person classes with limited accessibility for people who want to learn at their own pace, in their own language. Existing digital resources are static PDFs or simple websites that don't adapt to the learner's pace or provide interactive guidance — and the few apps that exist tend to gamify the sacred or fragment the experience across a dozen single-purpose tools.
Qué construí
Designed a multilingual learning experience with 5 AI companions
Built a complete formation platform supporting Portuguese, Spanish, English, and French. Each lesson includes structured content, interactive elements, and progress tracking, with five LLM-powered AI companions (on the Anthropic API) learners can talk to for thoughtful, grounded answers — the companions adapt their language and complexity to the learner's level and age mode.
Shipped audio ML and payments
Integrated audio and live pitch-detection ML for Gregorian-chant practice (audio never leaves the device), plus Stripe donations — in a single, solo-built Next.js + Firebase app backed by the Anthropic API. Ten content modules span catechism, daily liturgy, the Rosary, the Liturgy of the Hours, and more.
Shipped end-to-end, solo
From product strategy to UI design to backend architecture to deployment — every function was handled solo, with AI agents as the primary execution layer. Technically the most complex AI product in my set: multi-agent, audio ML, payments, and full internationalization, live at arauto-agape.web.app.
Números
4 languages — Portuguese, Spanish, English, French.
5 LLM-powered AI companions via the Anthropic API.
Live pitch-detection audio ML and Stripe donations.
Built and shipped end-to-end, solo, with AI agents as the primary executor.
Pantallas
¿Interesado en trabajar juntos?
Agendar call