A team that won the IMF’s Climate Innovation Challenge created the Multi-Level Carbon and Policy Analysis (MLCPA) tool to examine emission trends and countermeasures. The journey to achieving net zero is a long and difficult one. Because of the urgent need to reduce carbon emissions at a rate unprecedented in human history, it is crucial to identify structural breaks in emission patterns and comprehend the factors behind them. When dealing with climate data that is non-stationary, however, this might be challenging using conventional methods. To tackle this problem and aid users in locating and analysing such structural breaks, as well as to shed light on potentially game-changing strategies for emission reduction, a machine learning-based, four-pillar toolkit was developed. The link includes live demonstrations and case studies to show how the tool’s sectoral views may make climate policy design more focused and to educate the public on the economic intuition and policy implications of the toolbox’s results.