Paul Bergbusch

• Increased M&A deal flow – built a RAG app for company search • Accelerated due diligence – built an agentic AI app for investment research • Monitored and predicted deal-level fund performance – built a web app for simulation of leveraged buyouts • Reduced product recalls – built a computer vision web app POC for a food labelling company • Ran a computer vision project through POC and MVP at a subsea drone services company • Ran a LLM marketing project through POC at a pharmaceutical company

• Created new products, streamlined operations, and reduced manual errors – led a data science team developing machine learning models for trade settlement prediction, NLP-driven message classification, and customer support automation

• Reduced margin requirements for customers – managed a strategic risk project, engineered requirements and developed mathematical models for cross-margining of OTC and exchange-traded interest rate derivatives

• Launched a derivative valuation service (sold to Tullett Prebon in 2010) – hired and led teams of analysts and developers

• Improved the P&L of the equity derivatives desk – developed numerical models for pricing variance & volatility swaps

• Created new financial products – developed a Monte Carlo pricing engine and a calibration engine for exotic interest rate derivatives; built binomial/trinomial trees for pricing employee stock options and convertible bonds

• Optimized uplink/downlink data flow – architected, designed, and prototyped a multi-sensor satellite data simulator • Improved geo-location – developed and calibrated models of satellite motion, pointing, geo-location, and radiometry

• Found and measured the world’s rarest decay of a subatomic particle (kaon decaying into pion + neutrino + anti-neutrino) – analyzed terabytes of data to detect rare kaon decay patterns (1 in 10 billion) using PCA, DFA, and neural networks

Measurement of rare kaon decay

Measurement of radiative muon capture on nuclei