Showcase App

CaloIQ – The AI-powered nutrition PWA

CaloIQ is a modern progressive web app (PWA) for the intelligent capture and analysis of nutrition data. It combines advanced multimodal artificial intelligence with lightning-fast performance and a seamless experience across all devices.

Open CaloIQ app

Available at: caloiq.ideengeberhaus.de

The user workflow in five steps

How the app works

1
Secure registration
Users simply sign up with their email address and confirm their account via an automatic verification link.
2
Flexible capture
Meals are captured directly via smartphone photo, snapshot, or simple text input (e.g. “chicken breast with rice and broccoli”).
3
Automatic timestamps
Date and time are recorded automatically for each entry or can be adjusted manually.
4
AI nutrition analysis
Google Gemini AI recognizes the meal components within seconds and automatically calculates calories, protein, carbohydrates, and fat.
5
Real-time dashboard
The personal dashboard provides a clear overview of the remaining daily budget, the macronutrient distribution, and the history of past days.

Technical architecture & infrastructure

CaloIQ was built consistently on a modern, scalable, and future-proof tech stack:

Frontend & framework

Next.js 16 with React 19 and TypeScript. Styling is based on Tailwind CSS, while @ducanh2912/next-pwa handles offline caching and the app manifest.

AI engine

Integration of the Google Gemini API (@google/genai) for visual and textual semantic analysis of meals.

Hosting & cloud

Hosted on an Azure App Service (Linux) instance under the subdomain caloiq.ideengeberhaus.de with an automatically managed SSL/TLS certificate.

CI/CD pipeline

A tailored GitHub Actions workflow builds the application automatically on every code push. Using the Next.js standalone build mode, the app is packaged into extremely lean containers, reducing cold-start times on the cloud server to under one second.

AI-assisted development (AI Co-Engineering)

The app was developed through agile pair programming in close collaboration between human design and continuous AI engineering. Architecture decisions, pipeline optimizations, SSL bindings, and standalone deployments were realized in record time through AI-assisted iterations.

Privacy, authentication & data storage

Protecting sensitive user data and transparent data processing are top priorities at CaloIQ:

Backend & database

Supabase serves as the backend-as-a-service, built on a high-performance PostgreSQL database.

Secure authentication

User management is handled encrypted via Supabase Auth (@supabase/ssr). Authentication tokens are processed using strict HTTPS cookie standards.

Row Level Security (RLS)

All database tables are protected by Row Level Security. Each user has read and write access exclusively to their own nutrition data.

Encryption

All connections between browser, Azure cloud, Supabase, and the Gemini API are continuously end-to-end encrypted according to current TLS/SSL standards.

At a glance

Area	Technology / standard
Frontend	Next.js 16 (standalone output), React 19, Tailwind CSS
PWA standard	Service Worker, Manifest JSON, mobile native view
AI model	Google Gemini API
Database & auth	Supabase (PostgreSQL with Row Level Security)
Cloud hosting	Azure App Service (Linux Node.js environment)
Deployment	GitHub Actions CI/CD pipeline
Domain & security	Custom subdomain, automated Azure Managed Certificate (SNI SSL)

Try CaloIQ now

Experience how AI-powered nutrition tracking works in practice – right in your browser, with no installation.

Go to the CaloIQ app