Cook County Assessor's Office Open Source AVM (Automated Valuation Model) (github.com)

The Cook County Assessor's Office has open-sourced its Automated Valuation Models (AVMs) designed to estimate the assessed values of residential properties across the county. Their work includes two predictive models: one for single-family and multi-family homes (up to six units) and another specifically for condominiums. These models leverage historical data and property characteristics to provide more transparent and consistent property valuations, helping streamline tax assessments and appeals processes. This initiative is significant for the AI/ML community as it showcases the practical application of predictive modeling in public sector property assessment, offering a rare example of government-developed, open-source AI tools aimed at increasing fairness and transparency. By releasing not only the models but also the underlying data architecture, tax simulation tools, and comprehensive documentation, the Assessor’s Office provides a valuable resource for data scientists and policymakers interested in replicating or improving predictive valuation systems. Technically, the project encompasses robust ETL/ELT pipelines, SQL-based data lineage tracking, and an R package for simulating counterfactual property tax bills, enabling detailed analysis of how assessed values affect taxation. This blend of machine learning, public data, and open-source tooling underscores the growing role of AI in local government operations and property economics.