This week, Amazon Science announced that it has added automated reasoning to its research areas. The impact of automated reasoning can be seen in Amazon Web Services, where customers have access to features like IAM Access Analyzer and S3 Block Public Access. Amazon development teams are also integrating automated reasoning tools into their processes to improve the security, durability, availability, and quality of products. In this article, the author provides an introduction to automated reasoning for industry professionals. They explain how automated reasoning tools use mathematics to answer questions about programs and logic formulas. While the tools can be incredibly fast, they may sometimes respond with “Don’t know”. The goal of automated reasoning is to minimize these uncertain answers and provide more useful and powerful tools. Automated reasoning can be applied to both code and policies, and the tools help reason about control structure, what is eventually true, and what is always true in a program. While automated reasoning can significantly reduce the need for testing, it cannot completely replace it. Testing helps validate assumptions about compilers, runtimes, and other components of the computational stack. Automated reasoning tools sometimes return “Don’t know” or run forever, as certain problems are intractable. However, the challenge is to improve the tools to provide answers as often as possible
https://www.amazon.science/blog/a-gentle-introduction-to-automated-reasoning