Alexander Li Email & Phone Number

Austin, TX, US

San Francisco, US

• Developed and optimized code using Python, Java, C++, JavaScript, and HTML for training generative AI models, improving their precision in coding tasks.
• Engineered and fine-tuned prompts for AI models to generate precise and technical solutions for complex machine learning problems.

Chicago, Illinois, US

• Created applications to enable traders to look up and analyze trades.
• Used SQL queries in Python for data extraction and Apache Spark (PySpark) for analysis.
• Solely responsible for creating multiple Dash web applications from conception, to design, to integration and full production
• Automatically generated weekly trade metrics report in Databricks.

Chicago, Illinois, US

• Responsible for analyzing the largest factors behind gains and losses over a 10 minute time horizon based on theoretical valuation, and presenting the results.

Verona, WI, US

• Responsible for new product feature development, customer response and app issue resolution
• Developed a new client UI in Epic’s medication order app to associate medications with pharmacies.
• Debugged and fixed customer-reported issues using TypeScript (a superset of JavaScript) and C#.
• Facilitated communication between server and client regarding information about medications and parameters.

Austin, Texas - TX, US

• Data analysis of cryptocurrency transactions.
• Designing algorithms to detect anomalous cryptocurrency transactions.

New York, New York, US

• Implemented a variety of outlier detection algorithms to detect anomalies in trading data.
• Designed a RESTful API that allows users to input data in JSON format and runs a grid search to determine the optimal parameters and the models to use in an ensemble voting classifier.
• Serialized the models and stored them in an internal MongoDB database.
• Wrote unit tests on the algorithms and various different possible user inputs.
• Used Git to commit and keep track of changes.

Austin, Texas, US

• Reduced mean absolute error in house pricing models by 30-40% in out-sample testing set.
• Used various machine learning algorithms (such as Ridge, Lasso, Elastic Net, Random Forest, etc.), along with parameter settings, to quickly build up models and to compare model prediction accuracies.
• Applied traditional text analytics to analyze the sentiments embedded in listing comments.
• Wrote automated programs to find most common phrases in real estate brokers’ comments, and summarize average model errors for records with the phrases to extract valuation signals.

Austin, Texas, US

• Extracted large amount of data from zipped text files onto SQL database, standardized data fields and structured tables for efficient processing.
• Ran statistics for reverse mortgage borrowers and loan features, and evaluated the time trend in the program and among the borrowers.
• Compared the life expectancy between general population and reverse mortgage borrowers.
• Analyzed the important drivers for reverse mortgage terminations and built multivariate logistic regression model, all in Python.

Austin, Texas, US

• Processed tick-by-tick stock transactions to evaluate various technical trading strategies.
• Built Python codes to run Monte Carlo simulations, reconcile with valuation results using binomial lattice, and extract implied volatilities for real estate finance firms from equity option prices.
• Analyzed tax records for 70 million US single-family homes, to evaluate the time decay of land and improvement values across geographic regions.
• Ran multivariate logistic regressions to predict the time decay of property improvement values.

Master Of Science - Ms, Computer Science

Bachelor Of Science - Bs, Mathematics