Zhongyi Wu Email and Phone Number

Led a team of four to develop a drug intelligence public opinion monitoring and updating system, which developed from scratch, autonomously analyzes public sentiment and updates drug intelligence data, streamlining operations and improving decision-making processes.- Achieving a increase in efficiency compared to manual processes, Release 3 full-time employees to other more important jobs.- Applied for seven patents in AI-driven biomedicine applications.- Led daily text data cleaning and preprocessing efforts, enhancing data quality and facilitating downstream NLP tasks. Implemented noise removal, tokenization, and stop word removal techniques, resulting in a reduction in data processing time.- Spearheaded the development and optimization of text classifiers using TensorFlow and BERT, achieving F1 score improvement in accuracy for medical text classification tasks. Conducted drug name entity recognition (NER), improving recognition accuracy by Token-Level F1 Score through hyperparameter tuning and model fusion.- Orchestrated the integration of NLP models into practical applications, leveraging TensorFlow Serving for seamless deployment and performance optimization. Enhanced efficiency and usability of NLP-powered applications by F1 score.- Pioneered the integration of ChatGPT into medical industry prompts, refining prompts for enhanced information extraction from literature. Streamlined the process of extracting clinical results, new drugs, and targets from medical literature, accelerating knowledge discovery.- Conducted LLM model training and fine-tuning using pharmaceutical industry data, leveraging LLaMA 7B architecture. Achieved Human Evaluation, demonstrating the applicability and effectiveness of LLM models in pharmaceutical research and development.

Responsible for the research and development of artificial intelligence projects in the financial field of the company, responsible for the design and implementation of the company's credit investigation projects.- Spearheaded research and development initiatives for artificial intelligence projects within the financial domain, focusing on enhancing the company's credit investigation capabilities.- Led feature engineering efforts for financial quantitative trading, contributing to the development of robust quantitative stock selection factors.- Developed and implemented machine learning models for quantitative stock selection programs, leveraging a comprehensive library of over 900 factors. Utilized advanced algorithms such as SVM, Random Forest, Isolation Forest, XGBoost, and LSTM networks for automatic parameter optimization and iterative model refinement.- Achieved optimization of quantitative stock selection models through modification of objective functions and iterative optimization techniques- Demonstrated expertise in evaluating and selecting the most effective machine learning algorithms and methodologies to improve the accuracy and efficiency of stock selection processes.

Responsible for the implementation of on-site projects of the company's data analysis group and the establishment of enterprise credit information platform.- Led the construction of the company's corporate credit platform, including the design and implementation of credit scoring algorithms and models. Conducted comprehensive batch interface tests with cooperative companies to ensure seamless integration and functionality.- Spearheaded the development of the company's data warehouse, overseeing data cleaning and classification processes to ensure data integrity and usability. Utilized advanced statistical analysis techniques to explore the value of existing data and inform strategic decision-making.

- Participate in Business Intelligence (BI) program projects, conducting thorough analysis of report requirements and stakeholder needs to ensure alignment with business objectives.- Managed end-to-end data processing pipeline, including Extract, Transform, and Load (ETL) processes, to ensure accurate and timely delivery of data for reporting purposes.