AqualibriuM

Inspiration

Everyday fitness and wellness apps often operate in a silo. They track calories or log workouts, but they completely ignore a fundamental reality: physical exertion must adapt to individual recovery state and stress levels. Inspired by marine conservation and deep-sea exploration, we wanted to build something immersive, functional, and visually striking. AqualibriuM bridges the gap between structured personal health tracking and ecological mindfulness by tying human biometric telemetry directly to the health of a digital coral reef console.

What it does

AqualibriuM is a dark sci-fi MERN wellness platform that unifies four core pillars:

Workouts Hub: Exercise catalog search, muscle and category filters, and custom routines tracking sets, reps, weight, notes, and volume summaries.
Nutrition Hub: Foundation food catalog, macro tracking per 100 g, serving-aware meal building, and saved recovery meals.
Progress Hub: User goals, dated biometric measurements, history records, and live MongoDB persistence.
Analytics & AI Assistant ("Captain Aquali"): Real-time biometric tracking (stress index, recovery percentage, heart rate, and breathing rate) mapped to reef states (calm_abyss, active_reef, storm_tide) with adaptive, high-tech eco-coaching.

How we built it

Frontend: Built with React, Tailwind CSS, and Lucide React icons to construct an immersive, responsive dark-mode marine command console. Real-time synchronization is driven by custom browser events (aqualibrium:data-updated) to eliminate stale cache states across tabs and components.
Backend & Database: Powered by Node.js, Express, Mongoose, and MongoDB for robust data persistence of user meals, workouts, goals, and logs—backed by clean fallback structures for zero-friction local execution.
Hybrid AI Engine: Combines deterministic local rule assessments (evaluating risk levels and routing user intent) with Google Gemini 3.8 Flash to deliver schema-constrained, JSON-formatted generative coaching that stays completely in character as Captain Aquali.

Challenges we ran into

Balancing AI Persona & Safety: Keeping an AI coach immersive and engaging for a hackathon demo while removing rigid, repetitive medical disclaimers required carefully restructuring both backend system instructions (server/routes/assistant.js) and rule-based fallback generators (server/ai/captain_aquali.js).
Real-Time Dashboard Reactivity: Ensuring that deleting or adding metrics in the Progress Hub instantly updated aggregate analytics and visualizations across components required implementing a clean custom event dispatcher (notifyDataUpdated).

Accomplishments that we're proud of

Successfully executing an end-to-end hybrid AI architecture where deterministic metrics ground generative text into hyper-relevant, thematic eco-coaching.
Crafting a cohesive, production-grade dark sci-fi aesthetic ("AqualibriuM") that treats human wellness metrics like deep-sea telemetry.
Delivering a fully stable, reactive, and database-backed app ready for live presentation at Temple OwlHacks 2026.

What we learned

We gained deep experience designing hybrid AI pipelines that blend deterministic backend logic with generative LLM prompts, optimizing full-stack React-to-MongoDB data flow, and building highly thematic user interfaces that stand out in a hackathon environment.

What's next for AqualibriuM

Integrating wearable device APIs (like Apple Health or Google Fit) for live biometric streaming.
Expanding the digital reef simulation so that user consistency and recovery directly unlock visual upgrades and interactive ecosystem events on the dashboard.