Zhiqian Xu

Summer AI Lab Intern at n+1 Institute

Selected for a competitive OpenAI-funded program from 25+ teams; built an original AI-powered application with mentor guidance, later recognized as a top 3 project and presented to OpenAI leadership and at the UW showcase. Built a responsive web platform in React and TypeScript supporting 10+ interactive components, including natural language search, real-time stock charts, and personalized financial narratives powered by 7+ backend AI agents. Developed a 7-agent AI system for professional-grade financial analysis, with an LLM-based Portfolio Manager synthesizing multi-strategy signals into personalized reports with 95%+ ticker resolution accuracy.

Software Engineer Assistant – Mobile Software Development at UW–Madison: Department of Computer Science

Collaborated with course instructors to design mobile development assignments in Android Studio, aligning technical challenges with core learning objectives such as UI responsiveness, data persistence, and exception handling. Built comprehensive unit and integration test suites in Java and Kotlin using JUnit, and integrated them into an automated grading system, enabling accurate, scalable evaluation across 150+ student submissions per project cycle. Reduced more than 20 hours of weekly manual TA grading effort while significantly increasing feedback precision, consistency, and standardization, enhancing instructional efficiency and improving student learning outcomes.

Software Engineer – Capstone Project at Google

Drove development in an agile team with Google engineers to build an AI-powered SQL query optimization platform using Google Gemini 1.5 Pro, targeting performance improvements for MySQL and PostgreSQL databases. Led frontend development and architected backend components for query validation and integration with an AI-driven retrieval-augmented generation pipeline, enabling real-time syntax checking and query rewrite visualization. Designed core system architecture that achieved up to 86% reduction in execution time on TPC-H benchmarks and 91% on Star Schema Benchmark, delivering validated results in a live demo to Google SQL stakeholders.

Software Engineer Intern at Morgridge Institute for Research

Developed machine learning analysis tools to evaluate protein pathway reconstruction algorithms, improving analyzing efficiency by 80% in comparative studies by utilizing graph analysis and ensembling techniques. Implemented software solutions for data analysis by integrating container (Docker) and formal workflow managers (Snakemake), streamlining workflow automation and enhancing computational efficiency in bioinformatics pipelines. Collaborated with a multidisciplinary team of 10+ researchers and software developers from different fields, demonstrating strong teamwork and communication skills to align software development with scientific objectives.

Summer AI Lab Intern at n+1 Institute

Selected for a competitive OpenAI-funded program from 25+ teams; built an original AI-powered application with mentor guidance, later recognized as a top 3 project and presented to OpenAI leadership and UW showcase. Built a responsive web platform in React and TypeScript supporting 10+ interactive components—natural language search, real-time stock charts, and personalized financial narratives powered by 7+ backend AI agents. Developed a 7-agent AI system for professional-grade financial analysis, with an LLM-based Portfolio Manager synthesizing multi-strategy signals into personalized reports with 95%+ ticker resolution accuracy.

Software Engineer Assistant – Mobile Software Development at UW–Madison: Department of Computer Science

Collaborated with course instructors to design mobile development assignments in Android Studio, aligning technical challenges with core learning objectives such as UI responsiveness, data persistence, and exception handling. Built comprehensive unit and integration test suites in Java and Kotlin using JUnit, and integrated them into an automated grading system, enabling accurate, scalable evaluation across 150+ student submissions per project cycle. Reduced over 20 hours of weekly manual TA grading effort while significantly increasing feedback precision, consistency, and standardization—enhancing instructional efficiency and improving student learning outcomes.

Software Engineer – Capstone Project at Google

Drove development in an agile team with Google engineers to build an AI-powered SQL query optimization platform using Google Gemini 1.5 Pro, targeting performance improvements for MySQL and PostgreSQL databases. Led frontend development and architected backend components for query validation and integration with an AI-driven retrieval-augmented generation pipeline, enabling real-time syntax checking and query rewrite visualization. Designed core system architecture that achieved up to 86% reduction in execution time on TPC-H benchmarks and 91% on Star Schema Benchmark, delivering validated results in a live demo to Google SQL stakeholders.