Search for a command to run...
i build agentic AI and scalable risk platforms for institutional finance at Clearwater Analytics
my background spans financial engineering, machine learning, and systems development — from quant risk platforms to grounded LLM systems for clinical and research use
b.s. computer engineering, boston university · m.s. computer science, georgia tech
building institution agentic AI and scalable risk/analytics platforms for institutional investment management, joined through the acquisition of Beacon Platform
researched grounding LLMs in clinical applications using RAG, biomedical fine-tuning, and multimodal data to reduce hallucinations
built a CI/CD automation framework for Snowflake resource deployment using Harness and Liquibase
built an internal RAG system with Llama-3 to automate RFI responses and a financial risk modeling prototype for multi-asset stress testing
built an automated data quality framework integrating Great Expectations with Snowflake for validation and compliance across financial datasets
m.s. computer science, computing systems specialization
4.0/4.0 — GPU hardware and software, data and visual analytics, ML for trading
researching AI agents and LLM security under Dr. Vijay Madisetti (IEEE Fellow)
b.s. computer engineering, machine learning concentration
FIRST Robotics Presidential Scholar (half-tuition, top 0.5% of engineering admits), Order of the Engineer
advising on digital infrastructure and designing a custom LMS for an early education franchise network
mentored 40+ scouts through rank advancement and supported 10+ eagle scout projects as assistant scoutmaster
coached 40+ middle school students in robotics and autonomous programming through FLL competitions
top 0.5% of BU engineering admits, half-tuition merit scholarship with research funding
highest rank in scouting, awarded to less than 0.02% of scouts
1,617 of 90,000+ applicants nationwide
selected as 1 of 100 from ~7,000 applicants for the Citadel Summer Invitational
Autonomous paper-trading harnessed entirely through Claude Code — a council of analyst subagents debates, reaches a quorum, and trades. Educational, simulated paper account, no LLM API keys. Use at your own risk.
Reactive computation graph for Python. Cache results, track dependencies, only rerun what changed. SQLite persistence for sharing state across AI agent subprocesses.
Cross-session collective learning layer for Claude Code and Codex agents
Parallelized radix-2 FFT in C with OpenMP, applied to financial signal analysis of stock log returns
A self-writing AI intel wiki — Claude routines write twice-daily briefs to a markdown vault, with ntfy alerts, CI health checks, and a desktop reader (dashboard · graph · files).
This project focuses on automating the deployment and management of Snowflake resources using CI/CD pipelines, specifically leveraging Liquibase for database migrations and Harness for orchestration.