For the formal PDF version, download it here.

Jason Cusati

djjay@vt.edu GitHub: djjay0131 LinkedIn: jason-cusati
Download PDF

Professional Summary

PhD candidate in Computer Science (SE with AI/ML/MI) at Virginia Tech, focused on the empirical methodology of evaluating LLM behavior and accumulating verifiable safety claims. Two papers on evidence-based reasoning in GenAI tools and knowledge accumulation in software engineering research (ICSE 2026; arXiv preprint). Background combines hands-on ML/LLM application work --- LLM-assisted code generation, ML-driven incident prediction, behavioral evals --- with 20+ years of secure-software delivery (CSSLP-certified, recent mesh-VPN/2FA platform hardening at Virginia Tech's ARCADE Lab).

Professional Experience

2024 -- Present
Virginia Tech — Blacksburg, VA
Doctoral Researcher --- Software Engineering with AI/ML/MI
  • Researching LLM-curated knowledge graphs as a provenance substrate for evidence-based research progression --- studying how agent-orchestrated curation (Navigator / Extractor / Continuation) improves faithfulness of downstream LLM responses compared to LLM-generated KGs.
  • Investigating LLM-curated KGs as engineering-knowledge substrates (construction-domain material takeoff), with an emerging VVUQ layer for reproducibility of engineering-simulation claims.
  • Coauthored Verification, Validation, and Predictive Uncertainty Quantification of an Euler--Bernoulli Beam Finite-Difference Model (AOE/CS/ME 6444, Spring 2026; advisor Dr.~Christopher Roy; coauthor Cheng-Shun Chuang) --- planned extension into a reusable VVUQ framework for engineering simulation.
April 2026 -- May 2026
Virginia Tech — Blacksburg, VA
Graduate Research Assistant -- ARCADE Lab (PI: Dr. Afsari)
  • Designed and built cloud infrastructure for the Multi Robot System (MRS) for Remote Construction Progress Monitoring, including network architecture, mesh VPN, REST API, and Web UI.
2024--2025
Yoh, A Day & Zimmerman Company — Remote, SC
ADF Solution Architect -- Core Specialty Insurance
  • Led a small SDE team delivering V1 of an Insurance Data Ingestion Platform for a specialty carrier: extended Microsoft's Common Data Model into a domain schema for the carrier's data types; reduced time-to-ingest by 50%.
  • Used Claude as engineering partner to scaffold pipeline code and generate ADF source-to-target mappings; hand-implemented the more complex transformation and validation logic.
  • Architected and implemented an API to orchestrate Azure Data Factory pipelines and data flows.
2021--2023
Duck Creek Technologies — Columbia, SC
Senior Technical Architect -- OnDemand SaaS Operations
  • Developed Python algorithm to predict Sev 1 incidents from logs.
  • Coordinated in PEN testing and consulted on risk mitigation options for pre go-live security assessment.
Earlier employment history includes tenure as Principal Engineer -- Sustained Engineering, Blackbaud; Solution Architect -- Enterprise Professional Services, Blackbaud; Software Engineering Consultant, DJJayNet Consulting; Chief Architect, LimeLeap Solutions; Director of Professional Services, LimeLeap Solutions.

Education

2027 (exp.)
Ph.D., Computer Science (SE with AI/ML/MI), Virginia Tech, Blacksburg, VA
2026 (exp.)
M.S., Engineering, Virginia Tech, Blacksburg, VA
Master of Information Technology, Virginia Tech, Blacksburg, VA
B.S., Economics, Virginia Tech, Blacksburg, VA

Selected Projects & Research

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: Google Cloud (REST API + web apps), Tailscale mesh VPN for robot networking, Raspberry Pi edge nodes handling internal VPN routing and onboard video encoding, 2FA-hardened Web UI. Architecture completed and proven in a 5-week GRA appointment; 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.
Security Vulnerability Classifier — Repository scanner classifying code against CVE mappings.

Publications

Journal Articles

2026
Virginia Tech Transportation Safety Index (VTTSI) — Jason Cusati, Cheng-Shun Chuang . arXiv preprint arXiv:2606.21154
Link
2025
Exploring the Evidence-Based SE Beliefs of Generative AI Tools — Chris Brown, Jason Cusati . arXiv preprint arXiv:2407.13900
Link

Conference Proceedings

2026
From Papers to Progress: Rethinking Knowledge Accumulation in Software Engineering — Jason Cusati, Chris Brown . Proceedings of the IEEE/ACM 48th International Conference on Software Engineering (ICSE 2026)
Link DOI

Technical Skills

AI/ML

scikit-learn; TensorFlow; KNN; RNN; CNN; LSTM; LLM; SVM; Linear Regression; Random Forest; Naive Bayes; Feature Engineering

Languages

Python; C#; Java; JavaScript; Bash; PowerShell; T-SQL; C++; Node.js

Tools

Azure; AWS; Docker; FastAPI; Streamlit; PostgreSQL; SQL Server; VS Code; PyCharm; Wireshark

Platforms

Windows; Linux; IIS; Apache; Tomcat

Miscellaneous Experience

Awards and Achievements

May 2026
Phi Kappa Phi, inducted member --- national honor society recognizing academic excellence across all disciplines.
May 2026
Omicron Delta Kappa, inducted member --- national leadership honor society.
2008
Customer Service Award, LimeLeap Professional Services.

Certification

  • Certified Jira & Confluence Administrator
  • Certified Scrum Master
  • Certified Secure Software Lifecycle Professional (CSSLP)