BERDI is an AI-enabled system developed by the Canada Energy Regulator that makes it easier for Canadians to access and explore data from Environmental and Socio-economic Assessments (ESAs) submitted since 2003. The system uses machine learning to extract structured data and metadata from PDF documents, enabling targeted keyword searches and exploratory browsing of information related to the potential impacts of pipeline projects on the environment and on the health and well-being of Canadians.
The system is designed to serve both government employees and the general public, providing transparent access to environmental assessment data through an online portal. BERDI includes important safeguards to identify and exclude culturally sensitive Indigenous Knowledge from the dataset, ensuring that sacred and sensitive information is protected. The system does not process personal information about individuals.
This system is currently in production and available online at www.cer-rec-gc.ca/berdi (English) and www.cer-rec.gc.ca/ciber (French), as well as through the Open Science Data Portal. It represents the Government of Canada's commitment to making environmental impact data more accessible and searchable for informed public engagement in energy project reviews.
