lesswAIste

Client

Technische Hochschule Nürnberg

Role in project

UX/UI Design - Webdesign - Logo - Front-End Development - Back-End Development

Services

Figma
javaScript
Vision AI
Front-End Dev
Back-end Dev
Webdesign
Photography
Graphic Design
Logodesign
php
mySql
flask
python
html & css

Real World Project

lesswAIste - Bakery Waste Tool with Yolo11 & Prophet Forecastin

The Bakery Waste Optimization Tool is a fully developed web application designed to minimize food waste in bakery operations while providing actionable insights into production efficiency, ecological impact, and financial performance. Built for a bakery chain with six branches, the tool integrates advanced AI image recognition, predictive forecasting, and real-time data visualization to empower data-driven decisions. It processes production data for items such as croissants, pretzels, and Kaiser rolls, tracking leftovers and calculating optimal production quantities.

React: A JavaScript library for building user interfaces.
Create React App (CRA): A toolchain for bootstrapping React applications.
Tailwind CSS: A utility-first CSS framework for rapid UI development.
PostCSS: A tool for transforming CSS with JavaScript plugins.
Autoprefixer: A PostCSS plugin to parse CSS and add vendor prefixes.

Making sustainability visible

Key Features and Implementations

AI Model Completion (Preview Version):
A pre-trained YOLOv10 Vision AI model (upgraded to YOLO11) integrated with the API for image uploads, result storage in the database, and linkage to the forecasting model.
Database Application:
Built with real production data from six bakery branches, capturing detailed records of leftover items (e.g., croissants, pretzels, Kaiser rolls).
Client-Side Visualizations:
A web app displaying daily, weekly, monthly, and yearly losses alongside total production using Chart.js. Side-by-side charts compare losses vs. overall production.
Ecological and Financial Analytics:
Calculations for CO₂ emissions, water usage, electricity consumption, and labor costs—broken down by day, week, month, and year. Visualizations highlight the environmental and monetary impact of waste.
Predictive Forecasting:
Facebook Prophet model for forecasting optimal production quantities, incorporating regression for holidays and custom events.
PHP Server API:
Handles communication between the web app and database, with secure data querying and response handling.
Design and UI Enhancements:
- New logo development
- Time-intensive high-fidelity mockups in Adobe Illustrator using realistic database data for team evaluation
- Iterative design adjustments based on feedback
- Full frontend implementation (HTML & CSS)
- Tooltips for user guidance
- Dark mode for improved accessibility
Manual test data supports initial validation, with calculations extending to long-term projections.