cv

Lead Software Engineer

The MITRE Corporation

Leading the ATT&CK software team and conducting mechanistic interpretability research on MITRE’s Federal AI Sandbox.

• Bootstrapped MITRE’s interpretability research effort; led sprint developing in-house tooling for SAE training and automated feature analysis
• Contributed Kubernetes support and Azure inference integration to Neuronpedia (open-source interpretability platform)
• Building agent-based system for ML experiment and training run management
• Lead 10+ open-source ATT&CK tools used by thousands of organizations worldwide
• Authored the ATT&CK Data Model — first codified expression of the ATT&CK taxonomy as a TypeScript library

Senior Software Engineer

The MITRE Corporation

Software engineer on the MITRE ATT&CK team.

• Designed, deployed, and maintained the production ATT&CK TAXII 2.1 server (attack-taxii.mitre.org)
• Built REST APIs and open-source Python/TypeScript libraries for ATT&CK data distribution

Network Engineer

The MITRE Corporation

Infrastructure engineer for MITRE’s Bedford campus datacenter.

• Designed and productionized Cisco ACI network fabric
• Completed BS in Computer Science at Boston University while working full-time

Lead IT Systems Engineer

Neoscope

• Lead integration engineer for managed services provider
• End-to-end infrastructure refresh projects, from assessment through delivery

Master of Information and Data Science

University of California, Berkeley

Information and Data Science

• Focus: Generative AI, NLP, and mechanistic interpretability
• Research: Causal effects of fine-tuning on sparse autoencoder features
• Capstone: F1 Safety Car Prediction Engine — selected for Berkeley Summer 2025 Showcase

BS

Boston University

Computer Science

Unveiling the Black Box: Causal Inference and Feature Analysis in Fine-Tuned Language Models Using Sparse Autoencoders

UC Berkeley DATASCI 266

Examined the causal effects of fine-tuning on language model interpretability using sparse autoencoders and mechanistic interventions.