Resume

B.S., Computational Modeling and Data Economics · Computer Science · Blacksburg, VA

• Coursework: Data Structures and Algorithms, Advanced Programming in Python, Probability and Statistics, Machine Learning.

Associate AI Engineer, Part-Time · Platform Generative AI · Bethesda, MD

• Retained from the summer cohort into a part-time role on the platform team through the academic year, owning scoped features end to end, reviewing pull requests in Git, and shipping production changes against live Agile sprint commitments alongside full-time engineers.

• Java
• Spring Boot
• Git
• Agile
• React

Artificial Intelligence Engineer Intern · Platform Generative AI · Bethesda, MD

• Scaled a Java and Spring Boot backend from 36 to 250 classes and 6 to 105 REST endpoints across 5 modules, working in an Agile team through daily standups, sprint planning, peer code review, pair debugging sessions, and unit testing coached by senior engineers on the wider platform team.
• Redesigned service-to-service communication onto a shared inbound adapter in a ports-and-adapters architecture, cutting the code required to onboard a new service by 74% and eliminating transport logic duplicated independently across every module in the platform.
• Built an automated end-of-day financial close pipeline with deterministic balance checks that halt the run and flag the unaccounted discrepancy, plus a React operations console giving support engineers live service activity and a single-trace-ID audit log for every run processed.
• Deployed Docker services to AWS EKS through Harness CI/CD pipelines behind automated health and performance probes, backed by PostgreSQL retrieval indexed over 147 files and Redis caching that kept repeat lookups off the production database entirely.

• Java
• Spring Boot
• AWS EKS
• Docker
• Harness CI/CD
• PostgreSQL
• Redis
• React
• Agile

Research Intern, Data Analytics (Immuno-Oncology) · Washington, DC

• Analyzed flow cytometry data in Python (pandas, NumPy) and authored drafts of the peer-reviewed Biomedicines review below.

• Python
• pandas
• NumPy
• Flow cytometry
• Publication writing

kirodotdev/KiroCrew · Agent orchestration workspace · Verified on GitHub · Oct 8, 2026

• Authored 124 pull requests to AWS Kiro Crew: 80 merged, 2 open, and 42 closed without merge; GitHub credits 80 commits to the repository's contributor graph.
• Prevented silently truncated speech transcripts across the dashboard, Slack, and shared messaging channels by checking provider limits and refusing recordings whose duration could not be verified.
• Stabilized CI and hardened the runtime — replaced wall-clock timing in flaky async tests with event-based waits, fixed a Windows file-handling bug, attached an AppArmor security profile to the systemd service, and aligned UI copy across all 11 translated locales the product supports.

• Agent orchestration
• Speech transcription
• Accessibility
• CI
• AppArmor

AI portfolio and stock analysis platform

• Built and deployed a full-stack React/TypeScript and FastAPI investment-intelligence platform with a managed Postgres database, live market data feeds, Plaid portfolio sync for real account holdings, and scenario modeling tools built for everyday retail investors.
• Engineered an 8-stage backend workflow spanning data resolution, parameter estimation, probabilistic forecasting, backtesting, and portfolio simulation, with deterministic code computing every financial metric shown to end users on the investment dashboard itself.
• Added exact-cent rebalancing previews, cash-flow-aware performance history, and enriched holdings with cost basis, unrealized gains, and quote provenance; the latest CI run passes 206 unit and integration tests across the backend and frontend.
• Hardened production with rate limiting, request-ID logging, and Sentry instrumentation; ran a pilot with 50+ users whose simulated portfolios returned measurable modeled gains overall.

• React
• TypeScript
• FastAPI
• Vercel
• Render
• Neon Postgres
• Plaid

Object recognition PWA and Chrome extension

• Built and deployed a Next.js/TypeScript web app and companion Manifest V3 Chrome extension that identify objects fully on-device via MobileNetV2 running on ONNX Runtime, with lazy-loaded barcode and OCR recognition paths for lower-end devices.
• Containerized a Python/FastAPI inference service as a non-root Docker deployment, hardening the public API with payload validation, rate limiting, provider timeouts, bot protection, and structured request logging on every inference call made against the service.
• Modeled durable client state in versioned IndexedDB with schema validation and legacy migration, plus PostgreSQL-backed auth under row-level security; verified with Playwright end-to-end accessibility and layout tests across both mobile phone and desktop browser viewports.

• Next.js
• TypeScript
• ONNX Runtime
• FastAPI
• Docker
• PostgreSQL
• Playwright

Biomedicines, 14(2), 317 · Research conducted at Lombardi