x-ray

## Problem & Solution Statement Modern software is highly interconnected. A developer may change a single function, class, or API in one file, while the real impact can appear across multiple modules, services, or tests. Existing development workflows often rely on developers manually tracing dependencies and running broad test suites after a change. This makes it difficult to understand the potential impact of a change before it is made or merged. **X-Ray** is a web-based software change impact analyzer designed to make these hidden dependencies visible before they become problems. It acts like an “X-ray” for a codebase, helping developers understand **what could be affected before they change the code**. Developers interact with X-Ray through a web interface by uploading a project or connecting a repository. X-Ray analyzes the codebase and builds a structured dependency graph using deterministic static analysis. For supported code, it identifies files, classes, functions, imports, function calls, and relationships between components. When a developer selects a component or analyzes a Git change, X-Ray traces both direct and transitive dependencies to calculate the potential **blast radius** of that change. Instead of showing only a complex dependency graph, X-Ray provides multiple ways to understand the impact. Developers can isolate a component into a hierarchy view, inspect affected callers and callees, identify potentially affected tests, and analyze a Git diff to see how a proposed change propagates through the system. The key innovation is the combination of **deterministic software analysis with grounded AI assistance**. The analysis engine provides the evidence, while IBM watsonx.ai and Granite can interpret that evidence, explain potential risks in developer-friendly language, suggest a safer change approach, and generate targeted regression tests based on the affected code and identified impact paths. X-Ray is designed for software developers, engineering teams, code reviewers, and organizations working with complex or rapidly changing codebases. A typical workflow is: **Scan → Map → Select Change → Trace Impact → Ask “What If?” → Explain → Generate Tests → Verify** What makes X-Ray different is that it does not attempt to replace static analysis with AI or simply generate code from a prompt. It creates a **developer safety layer around code changes**, combining observable dependency evidence with AI-assisted reasoning and action. By making hidden dependencies and potential regression paths visible before a change is merged, X-Ray can reduce manual investigation, improve regression visibility, and help developers make more informed changes. Its architecture also provides a foundation for future GitHub pull-request integration, IDE support, and additional programming languages. **X-Ray: See the impact before you change the code.**