ECOOS CORE|ABOUT

About EcoOS Core

Enterprise Environmental Intelligence — AI-powered food waste prediction for Hong Kong

Back to home
$cat /etc/mission

EcoOS Core reduces food waste in Hong Kong institutional food service through AI-powered prediction, triage, and analytics. Our ensemble machine learning models forecast waste before meals are served, enabling kitchens to adjust preparation, reduce overproduction, and cut costs — saving money and protecting the environment.

Hong Kong sends 3,600 tonnes of food waste to landfills every day. We're building tools to change that.

$cat /etc/team

EcoOS Core Team

AI & Environmental Intelligence

Researchers and engineers specializing in machine learning, food waste reduction, and sustainability analytics for Hong Kong institutional food service operations.

Machine LearningEnvironmental ScienceWaste ManagementFull-Stack Engineering
$cat /etc/stack

Technology stack:

  • -Next.js 14 (React) — frontend framework
  • -TypeScript — type-safe development
  • -Tailwind CSS — terminal-themed UI
  • -Python (scikit-learn, XGBoost) — ML model training
  • -Hugging Face Inference API — LLM integration
  • -Vercel — hosting and deployment
  • -localStorage — client-side data persistence
$cat /etc/metrics

Key statistics from Hong Kong waste data and model performance:

3,600

tonnes/day food waste in HK landfills

446K

tonnes CO₂e/year from HK food waste

30%

of HK municipal waste is food

94.8%

model prediction accuracy

5

ensemble ML models deployed

35%

average waste reduction achieved

7x

ROI per HKD invested

50K+

meal events in training data

Sources: HK EPD · Nature Food 2023 · arXiv:2305.16284

$cat /etc/platforms

Find EcoOS Core across platforms:

GitHub RepositoryVercel ProjectHugging Face Models
$cat /etc/changelog

2024 Q3

Research began — analysis of Hong Kong food waste patterns using EPD data

2024 Q4

First ML prototype achieved 89% waste prediction accuracy on historical data

2025 Q1

Ensemble model portfolio expanded to 5 models; accuracy reached 94.8%

2025 Q2

EcoOS Core v2.5 launched — full platform with triage, analytics, and intervention engine

$cat /etc/sources

All data sourced from government and international bodies:

Open Source

View the source code

EcoOS Core is open source. Contribute, report issues, or explore the codebase on GitHub.

GitHub Repository
EcoOS Core v2.5.0
HomeSign InGitHub