Core Viewpoint - The introduction of the AI-driven workflow otto-SR significantly accelerates the process of systematic reviews in medical research, reducing the time from 12 years to just 48 hours, while outperforming human capabilities in various metrics [1][3][38]. Group 1: AI Workflow Development - The AI end-to-end workflow otto-SR was developed collaboratively by institutions such as the University of Toronto and Harvard Medical School [2]. - The system integrates GPT-4.1 and o3-mini for screening and data extraction, completing tasks traditionally requiring years in just two days [3][5]. Group 2: Performance Metrics - In benchmark tests, otto-SR achieved a sensitivity of 96.7% compared to human performance at 81.7%, and a specificity of 93.9% [5]. - The data extraction accuracy of otto-SR reached 93.1%, significantly higher than the 79.7% accuracy of human reviewers [22][24]. Group 3: Systematic Review Automation - The workflow automates the entire systematic review process, from initial literature retrieval to data analysis, allowing for human-AI collaboration [7]. - The screening agent utilizes GPT-4.1 for literature selection, achieving high sensitivity and specificity during both abstract and full-text screening phases [15][16]. Group 4: Practical Application and Results - In a practical application, otto-SR was able to identify 54 previously overlooked studies from a total of 146,276 citations, effectively doubling the number of relevant articles [26][27]. - The system's ability to quickly reproduce and update reviews allows for timely responses to new therapies and public health challenges [38][39].