Jingyi Li Email & Phone Number

New York, Ny, Us

• Implementing named entity recognition and question answering 🤗 model architectures based on PyTorch and fine-tuning them on fundamental equity sell-side research email corpus to develop an intelligent AI research analyst capable of distilling large corpus of information into actionable company insights and market trends.• Scoping out a platform co-pilot project with partner teams, gathering requirements and setting priorities for research and development with colleagues.• Built an optimization system for equity index trading from scratch leveraging convex optimization and model predictive control (MPC), preceded by the development of a trading volume prediction model, defined benchmarks and metrics to evaluate the effectiveness of the system and fine-tuned hyper-parameters to optimize for risk-adjusted returns in historical backtests, in collaboration with professors Stephen Boyd, Emmanuel Candès, Mykel Kochenderfer at Stanford University.

Washington, Dc, Washington, Dc, Us

• Co-managed and invested $580,000 of university endowment in U.S. traded global companies, conducted emerging markets’ capital markets research, analyzed investment environment and sector performance, issued bi-weekly report on German region.• Conducted fundamental valuation with DCF model on 3 companies in automotive industry, researched 10-K reports, implemented sensitivity analysis and Monte Carlo simulation in VBA, performed model validation, made recommendations to fund asset allocation and implementation strategy.

College Park, Maryland, Us

• Detected outliers from Equity’s Quotes data (20G+) by studying feature distributions and identifying abnormal quoting behaviors, cleansed data with distribution analysis and percentile winsorization using R data.table; • Created a mega database on WRDS cloud in UNIX operating system with 300+ equity features using SAS.

College Park, Maryland, Us

• Innovated a high-frequency trading strategy in U.S. Equity Market based on predictive modeling of one-minute ahead bid-ask midpoint with an R2 of 0.371, built with 17 significant variables using R Generalized Linear Model.• Researched last trade price and bid-ask midpoint. Identified last trade price as less predictable and a better measurement of current market via predictive modeling and statistical analysis.• Researched Corporate Bond market using TRACE data, visualized bond market trading activities, transaction costs for trading parties, analyzed dealers’ market power, trading behaviors and spread.

Miami, Florida, Us

• Built and deployed an R Shiny web app for ad-hoc real-time queries and end-to-end integration with other services on Azure. Constructed data pipelines to port local Salesforce and Banner application database extracts into Azure cloud via Informatica. Streamlined delivery of IP-Identification data for financial aid fraud investigation.• Created data ETL pipeline in SQL Server Management Studio, automated student aging report data migration and updates from Excel to SQL database in SAS.• Developed and deployed two interactive student recruitment performance KPI Tableau dashboards. Collaborated with business units, made recommendations for student conversion process and marketing campaigns.• Researched and analyzed impact of New Student Orientation on student retention rate using R, pulled and analyzed data from SQL database, delivered conclusion and recommendation to Orientation Committee.• Migrated team repos from local VCS to GitHub Enterprise, automated migration using GitHub Create API in bash script in UNIX operating system, provided training to three teams on GitHub migration demo.• Built tree-based binary classification model for word-picture matching using R RRF, achieved an AUC of 0.85.

Master'S Degree, Quantitative Finance

Bachelor’S Degree, French (Trade And Economics)

Study Abroad, Programme Grande École I, Economics

Education record