The Driver Transpiler treats software explanation as a compilation problem, emitting human language instead of machine code. This role involves static analysis to parse code from any language using tools like tree-sitter, building syntax trees and symbol table representations, and building templating engines for structured content generation with LLMs. You will improve scalability and robustness, contribute to the compiler architecture, and own language specialization end-to-end. You'll advance the design and implementation of language specialization and contribute to the internal AutoDocs engine, expanding the transpiler's capabilities and introducing new optimizations. You will communicate effectively with backend engineers to deploy transpiler updates. You’ll work on technology at the intersection of language theory, compiler design, and generative AI building systems that expand how both humans and machines understand code.