Over the last year, we’ve been building Zeno, an AI-driven personal growth app. Here’s a quick look behind the scenes at our tech decisions and challenges faced along the way:
1. Modular by design
We structured Zeno around 9 "Growth Islands" (habits, motivation, relationships, etc.). This modular setup made it easier to deliver personalized experiences without overwhelming users.
2. AI that adapts in real-time
Instead of static recommendations, we integrated an adaptive AI system using GPT-4o with a RAG architecture (Retrieval-Augmented Generation), leveraging Pinecone for deep contextual insights.
3. UX is everything
Our biggest hurdle: making AI interactions feel human. We redesigned our chat interface multiple times (finally settling on Remix.run + Framer Motion) for fluid, intuitive interactions.
4. Global from day one
We added multilingual support early, using a scalable i18n pipeline. This opened new markets instantly, proving a small investment upfront can yield big returns.
Building Zeno taught us the importance of flexibility, empathy in UX, and early internationalization.
Happy to share more—any questions on our tech stack, architecture decisions, or product choices?