Benjamin Montesinos

Built a Java-based CLI tool for analyzing Excel trade journals, processing 800+ historical trades and computing key performance metrics such as win rate, expectancy, profit factor, and max drawdown. Structured project into data ingestion, analytics, and reporting components to improve maintainability and extensibility. Implemented exportable reports in JSON/Markdown and added optional LLM summary features.

Evaluated and ranked AI-generated code responses for correctness, instruction-following, edge-case handling, and overall software engineering quality across coding and debugging tasks. Wrote reference solutions and technical explanations to support model training in areas such as code generation, code completion, refactoring, and problem solving. Identified subtle logical flaws and inefficient implementations in model outputs.

Final project for course where two public data sets were analyzed to examine relationships between substance use, personality traits, and socio-economic factors, then built a Java-based CLI tool to present findings with structure. Collaborated in a team using Git-based workflows to integrate components, review code, and iteratively debug and improve the system.