myclimate – SmartPCF – AI-based CO₂ Accounting

Challenge

Companies want to assess the carbon footprint of their product portfolios, but manual analysis is too time-consuming. With thousands of products, individual calculation is not practical.

Solution

In close collaboration with myclimate:

• Django platform with intuitive web interface for product uploads
• ML models (Random Forest) for CO₂ estimates based on category, price, weight, and material
• Combination with comprehensive life cycle assessment data for realistic estimates
• Dashboards for portfolio analysis and emission hotspot detection
• Continuous retraining and validation of models

Our Contribution

• Technical architecture and platform development with Django/Python
• Development of intuitive web interface and scalable backend
• Integration of ML models (Random Forest, Generalized Random Forest)
• Development of dashboards and reports for portfolio analysis
• Operationalization and production operations with continuous retraining
• Ensuring stable, performant calculations for large product catalogs

Technologies

Results

Companies can now assess entire product catalogs in hours instead of months. The platform automatically identifies emission hotspots and prioritizes reduction measures – for more sustainable decisions with less effort.