Liming Nihao developed a dynamic script-based rescorer for the elastic/elasticsearch repository, enabling search scoring to be driven by customizable scripts that leverage document attributes such as num_likes and num_reviews. This feature enhanced the flexibility and personalization of search results by integrating extensible scoring logic directly into the Elasticsearch pipeline using Java and YAML. Later, Liming focused on backend stability in apache/solr, resolving a concurrency bug that caused ArrayStoreExceptions during multi-threaded reranking and sorting. Through targeted unit testing and backend development, Liming improved the reliability of high-concurrency ranking workflows, demonstrating depth in search algorithms and large-scale open source collaboration.