uflow

I was inspired by the universal struggle every student faces: the "paralysis of choice." We wake up knowing we need to study and build a career, but the sheer number of tasks, deadlines, and job applications is overwhelming. I wanted to build a system that doesn't just list tasks, but actually thinks with you. The concept of "Flow State"—that perfect zone of focus—became the core identity. I wanted to create a digital environment that feels liquid, adaptive, and calm, helping students slip into that zone effortlessly.

How we built it UFlow is a sophisticated Hybrid AI Architecture built to bridge the gap between structured data and creative intelligence. The Core: We built the frontend with Nuxt 3 and TailwindCSS to achieve a blazing-fast, "app-like" feel. The design system ** The Brain**: The backend is powered by Supabase for real-time data and authentication. However, the true intelligence comes from our dedicated Python Microservice (FastAPI). This service acts as the "Cortex", using advanced LLMs (via OpenRouter/Gemini) to analyze student moods and generate personalized daily schedules. The Agent: We took it a step further by integrating N8N automation. When a company accepts a student, it's not just a database update—an autonomous AI agent wakes up, schedules the interview on Google Calendar, and sends a personalized email invite, removing the administrative burden entirely. Challenges we ran into The biggest challenge was orchestrating the communication between the Nuxt server and the Python AI microservice while maintaining a seamless user experience.

Authentication: Securing the bridge between the client, the Nuxt server, and the Python service required a robust JWT handling strategy to ensure student data remained private.

Built With

• nuxt
• openrouter
• python
• tailwindcss
• vue