PhD Student · Federated Learning and Agentic AI
I build federated learning frameworks and multi-agent development systems. My research sits at the intersection of trustworthy distributed ML, human-on-the-loop AI agents, and developer experience for AI-assisted software.
Computing and Information Sciences · Rochester Institute of Technology
Research on coding agents that learn each organization's conventions and pool what generalizes, while each organization decides what may leave. Advisor: Dr. Leon Reznik. Also an NSF AWARE-AI National Research Traineeship (NRT) trainee in the AI Software track.
Federated Learning · Agentic AI · Code Review · LoRA
SCADS · Summer Conference on Applied Data Science · NC State · Laboratory for Analytic Sciences
Selected for a competitive summer research cohort applying data science to analytic sensemaking problems. Built agentic sensemaking and governed MCP orchestration prototypes and a motion-first tracking pipeline for camouflaged targets in video.
Claude Agent SDK · MCP · Agentic AI · Video Object Tracking
Software Architecture · Rochester Institute of Technology
Mentored students through design patterns, architectural styles, and quality attributes. Led a full class session during the instructor's absence.
Software Architecture · Design Patterns · Mentoring · Code Review · Teaching
Laboratory of Data Quality and Intelligent Security (LDQIS) · Rochester Institute of Technology
Built InteFL's pytest suite (80%+ coverage, run in GitHub Actions CI), its attack and defense tooling, and its Hugging Face text-dataset loaders. Built the lab's new website.
Federated Learning · pytest · GitHub Actions · Hugging Face · Astro
Software Engineering · Rochester Institute of Technology
GPA: 4.0/4.0. Capstone: enhanced the InteFL federated learning execution framework under Dr. Leon Reznik.
Machine Learning · Distributed Systems · Cloud Computing · Software Architecture · Python
Distribution Management Systems · Lowe's Home Improvement
Cut daily error logs by 150,000 across RDCs. Built full-stack automation tooling. Reduced onboarding time by 80%.
Java · SQL · Spring Boot · Full-Stack · Agile
Development Team · Phiquest
Built a blockchain-based document-authenticator Chrome extension in Node.js. Wireframed web-app surfaces with cross-functional teams.
JavaScript · Node.js · UX Design · Blockchain · Figma
Computer Engineering · University of South Florida
GPA: 3.06/4.0. Capstone: shipped the PhiQuest blockchain-explorer document-certification system with Allen Cheng and Ethan Wuitschick.
C · C++ · Embedded Systems · Digital Logic · Microcontrollers
A coordinated family of repos exploring agentic AI and federated learning from complementary angles. Each is independently invocable.
Multi-agent software-development forge: maidens-as-specialists over A2A, MCP sidecars, transparent human-on-the-loop.
Federated LoRA fine-tuning testbed on the current Flower release, with OpenTelemetry traces and metrics for every round and client.
Federated personalization testbed with security-labeled synthetic data. Measures what a personal/shared boundary costs and how much leaks past it.
Federated learning research toolkit with a Rust core, Python bindings (maturin), PyTorch and Hugging Face model support, and an MCP server interface.
Public website for the Laboratory of Data Quality and Intelligent Security at RIT.
Claude Code plugin marketplace: repository maintenance, adversarial code review, LaTeX and slide builds, and bibliography verification.
Agentic issue triage on Claude Opus and the Anthropic Message Batches API: classifies, prioritizes, and routes incoming reports at batch scale.
SCADS 2026 Technical Report (pp. 7-34), Laboratory for Analytic Sciences, NC State University · 2026
A Model Context Protocol server that lets an agentic harness propose queries while the server enforces policy, runs read-only connector operations, records provenance, and builds cited reports.
Team · RIT AWARE-AI · Oct 2026
Myo Lever Controller: drives a simulated Clearpath Jackal from one forearm.
Talk · Rochester Security Summit 2026 · October 2026
Talk by Dmitrii Korobeinikov, AJ Barea, and Leon Reznik on a decision framework for federated learning applications that detect and respond to attacks, balance security overhead against model performance, and adapt to changing conditions.
IEEE Intelligent Systems · 2026
We argue that metacognition should be the guiding principle for designing and deployment of Federated Learning (FL) applications.
Solo · HackerRank · May 2026
Multi-domain support-triage agent: routes tickets across HackerRank, Claude, and Visa corpuses with corpus-grounded responses, batch processing through the Message Batches API, and prompt-injection defense.
Poster · NE Agents Day 2026 · May 8, 2026
Reframes multi-agent coding as an interpretability problem via three pillars: Monitor (OpenTelemetry / Jaeger / Prometheus observability), Communicate (agents request clarification on ambiguous requirements), and Control (bounded repair loops for deliberate recovery).
IJCNN 2026 (IEEE WCCI, Maastricht) · 2026
Although Federated Learning (FL) enhances clients' security and privacy by retaining data locally, the decentralized nature of this paradigm exposes it to diverse malicious actions as well as technological and reliability issues that can degrade data quality, hinder convergence, and reduce global model accuracy.
Team · RIT AWARE-AI · Feb 2026
Cosmic Ritual: BioRadio EMG gestures drive a real-time chord engine. Built the gesture classifier, MIDI engine, GUI music mode, and Zensical docs site.
Poster · CHAI Research Poster Session 2026 · April 15, 2026
Examines how embodiment, voice, and Ren'Py visual-novel framing affect how users supervise and redirect multi-agent software.
Email is the fastest way to reach me.