This research introduces a hybrid approach combining LLMs and static analysis to automatically translate C code to safer, idiomatic Rust code with correctness guarantees.

• Integrates static analysis with LLMs to preserve program semantics
• Uses a multi-step translation pipeline for handling complex C features
• Achieves more reliable and idiomatic code translation than traditional methods
• Significantly reduces manual effort while maintaining correctness

This innovation addresses critical engineering challenges by automating the transition to memory-safe languages without sacrificing performance, helping organizations modernize legacy codebases with greater confidence and reduced security risks.

LLM-Driven Multi-step Translation from C to Rust using Static Analysis