From the course: Accelerate DevOps and Software Development with AI: Modern Tools and Workflows for Enhanced Software Delivery

Unlock this course with a free trial

Join today to access over 25,200 courses taught by industry experts.

Case studies: AI wins and fails

Case studies: AI wins and fails

From the course: Accelerate DevOps and Software Development with AI: Modern Tools and Workflows for Enhanced Software Delivery

Start my 1-month free trial Buy for my team

Case studies: AI wins and fails

“

- Now that we've played around with a couple of examples, let's shift the focus to stories from the industry. We'll begin with a clear success story. Morgan Stanley's DevGen.AI Morgan Stanley had millions of lines of legacy code, hard to read, hard to maintain. Modernizing this code would be a headache. Instead of asking developers to decode it manually, they used a custom GPT-based system to automatically generate English documentation from the code, making the modernization a lot easier. Since it's launch, it's been used to review over 9 million lines of code, saving developers 280,000 hours of manual work. The takeaway here is clear, AI worked because it was applied to a well-defined high friction problem. Understanding old code, not to something critical like writing production logic. It was a translator, not a decision maker. Now let's look at a high profile fail, Replit's AI assistant deleting a database. Replit's…

Contents