Projects
Multi Robot System — Cloud + Edge Architecture (VT ARCADE Lab)
Designed and built the cloud-and-edge architecture for the ARCADE Lab's Multi Robot System for remote construction-progress monitoring during a 5-week GRA appointment (April--May 2026): Google Cloud (REST API + web apps), Tailscale mesh VPN for robot networking, Raspberry Pi edge nodes handling internal VPN routing and onboard video encoding, Terraform-based infrastructure-as-code and CI/CD pipelines, and a 2FA-hardened Web UI. Claude used as engineering partner throughout.
VTTSI — LLM-Adaptive Intersection Safety Scoring
Distributed rewrite of a CS6604 traffic-safety application: provenance-tracked evidence ingestion (NWS weather, VDOT/511 cameras) → deterministic RT-SI + MCDM baselines → Claude Opus adaptive adjustment with grounding guardrails and deterministic fallback → honest-null ablation eval (Spearman / Pearson). Python ports-and-adapters microservices; live-Claude integration wired; ~97% offline coverage.
Agentic Knowledge Graphs for Research Progression
Orchestrated agent system (Navigator / Extractor / Continuation) building an LLM-curated research KG --- problems, assumptions, constraints, evidence spans, datasets, metrics --- as the provenance substrate for downstream research-assistant responses. Denario agent framework; Python / FastAPI; staging deployed on Cloud Run. Approaching experimental comparison against LLM-generated KGs.
Construction-AI: LLM- and KG-backed Material Takeoff
Web app parsing architectural drawings (DWG / DXF / PDF), extracting geometry via PyTorch / OpenCV / EasyOCR, producing JSON material takeoffs with OR-Tools cut-list optimization backed by a Neo4j LLM-curated KG. Phase 1 MVP deployed on Cloud Run; Phase 5 in progress adds a VVUQ (verification / validation / uncertainty quantification) layer for engineering-simulation reproducibility.
VVUQ of an Euler--Bernoulli Beam Finite-Difference Model
Verification, Validation, and Predictive Uncertainty Quantification of a finite-difference beam solver for residential LVL header design (AOE/CS/ME 6444 final paper, Spring 2026; advisor Dr.\ Christopher Roy; coauthor Cheng-Shun Chuang). Planned extension into a reusable VVUQ framework.
LLM Coding Assistant
Python, Streamlit, OpenAI API; FastAPI backend; agentmon design.
Security Vulnerability Classifier
Repository scanner classifying code against CVE mappings.
Sentiment Analysis of Yelp Reviews
Python, Pandas, NLTK; topic modeling with LDA.
Predicting NFL Player Performance
Time-series modeling using ML regressors.
LLM-Assisted Data Importer (Insurance)
Internal prototype (2024--2026) of an LLM-driven data-import pipeline for the insurance domain, using Claude for structured extraction, mapping, and validation of heterogeneous carrier input sources.
Common Insurance Schema Framework
Internal prototype (2024--2026) of a common-schema framework for insurance data, built with Claude assistance to reconcile carrier-specific data models into a shared canonical form.
KGIS Documentation
What KGIS is, why it exists, its architecture, and how to use it: the deterministic, evidence-first knowledge-graph ingestion platform.