Even those fortunate to understand how to make compliance tools using artificial intelligence (AI) still find the process intimidating. Every company’s needs are different and building an AI solution will likely involve roadblocks, challenges, trial and error. While researching our series on AI solutions for compliance, we spoke with experts at AB InBev, Microsoft and Google about how their companies approached building tools using machine learning. Here, we walk through the individualized process for each company and offer practical advice for getting a project off the ground. See our three-part series on AI for compliance: “Foundations” (Nov. 11, 2020); “Building a Model” (Dec. 16, 2020); and Five Workarounds for Asymmetric Data Sets (Mar. 10, 2021).