I Built an App That Picks Films Based on Your Mood

Mood Movie Recs AI, pushed in February 2025, started as a playful idea and turned into a lesson in recommendation logic. We all ask the same question on Friday night — what should I watch? — and I wanted an app that answered with more personality than a random scroll. Built in Kotlin, the project combined user mood inputs with AI-driven suggestions, forcing me to think about how interfaces can feel conversational without becoming gimmicky. I cared about the experience: quick prompts, clear results, and enough transparency that users understood why a film appeared. The last push in May 2025 came after UI refinements I had been postponing, which taught me that launch and polish are different verbs. This repository sits between my NLP work and my GenAI sprint — a bridge project that kept me honest about mobile development while my head was drifting toward LLMs. It also reminded me that not every repo needs to be enterprise-grade to be valuable. Some repos teach taste, and taste is what makes larger systems feel human when they finally ship. I still think of it as the project that kept my feet on the ground while my ambitions grew taller. Recommendation systems look easy until mood, genre, and novelty fight in the same UI. This app taught me to negotiate those trade-offs in Kotlin before I negotiated them in production-minded web stacks. Mobile reminded me that users do not care about your stack; they care whether the app feels thoughtful in under ten seconds.