Isaac L.

Led projects in Applied Science and Engineering for AI. Some highlights: - Engineered a full-stack automated LLM red-teaming platform supporting single-/multi-turn, contextualized, and multilingual jailbreaks, observability and detailed performance metrics. Jailbreak GPT-4o within 30 seconds, with up to ~90% attack success rates. Scaled to 55k+ queries and rolling out for productionisation. - Built AI systems for cybersecurity: - Automated Pentesting: Designed core agent modules, frontend, backend-frontend architecture for real-time event streaming via SSE, S3/DynamoDB integrations - Adversarial Tooling: Developed agentic workflows for offensive research and implemented an MCP-based orchestration layer for isolated, ephemeral VM sandboxes with real-time human-in-the-loop oversight via websocket. - Developed a 2-stage RL pipeline for full LLM uncensoring and a SFT+RL pipeline for effective localized safety alignment (Best Paper@AAAI 2026 LM4UC Workshop; https://openreview.net/forum?id=x9KuyCwksR). Built modular, serverless pipelines for end-to-end LLM finetuning, evaluation and experiment tracking on GCP Vertex AI, scaling to 100+ experiments.

- Built real-time yield curve on serverless architecture for monitoring of municipal bond markets - Experiments in feature engineering and custom model architectures to improve accuracy, enforce non-crossing bid-ask predictions, etc - Built serverless ML pipelines for repeatable and scalable automated training

Deep factor analysis for mutual funds using layerwise-relevance propagation for DNNs and LSTMs

Real estate valuation in high-dimensional, high-cardinality data applying spatial econometrics, deep learning and explainable ML

Analyzed heterogeneous treatment responses to income tax rebates among US households using causal forests, an extension of random forests to causal inference with defined asymptotic properties

Fall 2020: Applied Deep Learning, Business Analytics, Sports Analytics, Optimization Models and Methods, Probability and Simulation Spring 2021: Asset Management, Machine Learning for Operations Research and Financial Engineering, Deep Learning for Operations Research and Financial Engineering, Stochastic Models, Operations Consulting Summer 2021: Marketing Analytics, AI Applications in Finance

Relevant Coursework: Applied Statistics, Quantitative Economics and Econometrics, Accounting for Business, Microeconometrics, Econometrics for Macroeconomics and Finance, Advanced Linear Algebra, Computational Methods for Economics, Computational Methods and Programming, Independent Research Paper