Sustainability Impact Assessment

A major global e-commerce company developed and deployed Flamingo, an AI-based algorithm using natural language processing to automate environmental impact factor matching for product carbon footprinting. Published in the ACM Journal on Computing and Sustainable Societies in August 2023, the research demonstrated 75% precision on a 664-product dataset. In practical application, the algorithm reduced the time scientists spent mapping 15,000 products from approximately one month to several hours, enabling the company to scale carbon footprint calculations across its vast product catalog. The algorithm has been made available as an open-source tool for other organizations to use.

A global multicategory food company, as documented in the 2024 BCG and CO2 AI Carbon Survey, used AI to match more than 115,000 products to individual emissions factors. With 98% of the company's carbon footprint residing in Scope 3, the AI-driven approach significantly automated the emissions measurement process, improving both accuracy and efficiency. Separately, a global supplier of ingredients for fragrances, flavorings, and cosmetic active ingredients has been using AI-powered carbon accounting to compute product carbon footprints across 10,000 raw materials and 90 production sites, as reported by CO2 AI in 2024. A global consumer goods company also used AI-enabled product lifecycle assessment to cut Scope 3 emissions associated with one of its air freshener brands by reducing emissions in transportation, manufacturing, and raw material sourcing.