This project develops a concise domain-specific language and AI assistant for designing and validating financial instruments. The system connects developers in financial contexts–quant developers, risk/trading teams, and integration engineers–to pricing engines, market data, booking, and reporting systems. It aims to compile natural language requests into specifications, including payoffs, schedules, and valuation settings. This pipeline aims to produce shorter time-to-market, fewer back-and-forths, and decisions that meet regulatory restrictions in a range of financial settings. The DSL standardizes how ideas are specified and reviewed, while the assistant surfaces edge cases, explains trade-offs, and produces answers and rationale that teams can trust.

This SuperUROP gives me the opportunity to bridge the gap between CS, machine learning, and finance, and study how modern ML can promote problem-solving in the fintech domain. Having experience working on models and designing systems in settings from academia to market-making firms and exchanges, I am interested in translating well-founded techniques into fintech capabilities and learning to frame research questions into tools that can make an impact in a range of financial contexts.