C++ Developer

Main Objective: To enhance debugging efficiency for large language model algorithms and strengthen infrastructure support for AI-driven systems. My Contributions: Designed and implemented scripting tools to streamline algorithm debugging workflows, including online log filtering, multi-agent request dispatching, result comparison, and diff-based visualization of discrepancies. Built a space-efficient Trie-based tokenizer using double-array compression, supporting maximum-probability segmentation and pattern matching; extended its capability with an Aho-Corasick (AC) automaton to enable user-defined vocabulary detection. Leveraged the AC automaton for real-time keyword scanning and implemented robust filters for URLs and quoted text in input streams. Contributed to the development of an inverted index system for an internal AI search engine, improving retrieval accuracy and latency. Enhanced system reliability by reducing manual intervention through automation, and improved content processing robustness in edge cases such as malformed inputs or emergency aborts.