We built a "video player" for Python scripts that tracks runtime variables and catches silent ML bug

Have you ever opened a Python script from a teammate, classmate, or tutorial and wondered, "What is this actually doing?" Reading someone else’s code line by line is exhausting, and asking generic AI chat tools often gives vague summaries that miss what really happens when the code runs. TraceLens changes that by turning code into an interactive, visual journey. Think of it as a video player for your software: instead of guessing, you can hit play, pause, step forward, or rewind to watch your program run frame by frame. Here is what TraceLens does for everyday developers: Watch Your Code in Motion (LogicLens): Step through your script at your own pace. TraceLens brings loops to life so you can inspect each iteration, clearly highlights why an if/else decision took a specific branch, and shows a live, color-coded board of your variables as they are created, modified, or cleaned up. Know Exactly Where to Add New Features: One of the most intimidating parts of touching inherited code is the fear of breaking something. TraceLens automatically discovers Safe Insertion Points—friendly star markers in the code editor showing you the exact spots where data is stable and ready for you to safely hook in your own new logic. Your AI Pair-Programming Partner with IBM Bob: Powered by IBM Bob, TraceLens gives you a friendly guide sitting right beside you. At each step, Bob explains the author's true intent and design decisions in plain English. When you are done, Bob generates a clear teammate handoff summary so anyone can pick up the project without confusion. Catch Sneaky Data Science Traps (ModelLens): If you build AI or data pipelines, silent bugs (like accidentally leaking test data before training or using heavily imbalanced datasets) can ruin your model without ever throwing an error message. TraceLens flags these hidden traps instantly and provides one-click fixes to keep your models healthy.