cv

Software Developer

Medical Informatics Engineering

Rebuilt a production healthcare compliance platform's backend from Java 21/Spring Boot to modular TypeScript as sole developer, with zero breaking changes for API clients.

• Implemented HL7 FHIR R4 EHR interoperability and regulatory exports for eight CMS quality measures, validated against 410 official MADiE test cases, and reported a confirmed reference-engine defect.
• Designed a resumable batch pipeline over 70,000 PostgreSQL records with bounded memory and per-record failure isolation.
• Shipped guardrailed AI assistants and read-only MCP tools, backed by 785 automated tests in CI and 43 architecture decision records.

Product Associate

Riccle

Analyzed customer, transaction, and competitor data to shape positioning and channel strategy.

• Automated go-to-market workflows that cut six hours per week of manual reconciliation.

Graduate Researcher, Machine Learning Evaluation

Purdue University Fort Wayne

Evaluated how adaptation methods and measurement choices affect jailbreak robustness for open-weight language models.

• Automated SLURM pipelines fine-tuning 25 model configurations across five families with LoRA, QLoRA, full fine-tuning, and 4-bit quantization on A100 and A30 GPUs.
• Ran an evaluation suite of nearly 600 measurements, exposing a 45.7% false-positive rate in a widely used automated evaluator versus 8.6% for a calibrated alternative.
• Published three first-author works, including two SemEval-2026 system papers in the ACL Anthology.

Software Development Intern

Medical Informatics Engineering

Built a containerized RAG platform for citation-backed search over enterprise documents.

• Architected hybrid search and reranking with Node.js, Docker, OpenSearch, Redis, and PostgreSQL over 3,000+ documents at 85%+ context relevance.
• Exposed REST, MCP, and SSE interfaces with structured citations, JWT/API-key authentication, user-scoped security filters, and OpenTelemetry observability.

M.S.

Purdue University Fort Wayne

Computer Science

• Master's thesis: Do Efficient Adaptations Reduce Safety? Jailbreak Robustness of PEFT vs. Full Fine-Tuning on Consumer-Accessible LLMs
• Thesis advisor: Dr. Jonathan Rusert

B.S.

Lahore University of Management Sciences

Computer Science

Do Efficient Adaptations Reduce Safety? Jailbreak Robustness of PEFT vs. Full Fine-Tuning on Consumer-Accessible LLMs

Purdue University Fort Wayne master's thesis

Master's thesis on adaptation methods, jailbreak robustness, evaluator disagreement, and safety-efficiency trade-offs in consumer-accessible LLMs.

PFW Task 8 at SemEval-2026 Task 8: Lightweight Tri-Fusion Retrieval with Prompt-Engineered Faithful Generation for Multi-Turn RAG

Proceedings of the 20th International Workshop on Semantic Evaluation

First-author MTRAGEval system paper; the system ranked 6/26 on Task B.

PFW at SemEval-2026 Task 6: Multi-Seed DeBERTa Ensembles for Political Response Clarity and Evasion Classification

Proceedings of the 20th International Workshop on Semantic Evaluation

First-author system paper for CLARITY; the system ranked 18/41 on clarity and 12/33 on evasion.

Your Judge Is a Confound: Evaluator and Attack Choice Distort Jailbreak-Safety Measurement for Fine-Tuned LLMs

AdvML-Frontiers × CoTMA Workshop, COLM 2026

Accepted first-author paper on how evaluator and attack choice distort jailbreak-safety measurement for fine-tuned LLMs.

JobOps Copilot

Deployed AI job-operations platform on Azure with a Next.js frontend, Express/TypeScript API, and FastAPI agent service.

• LangGraph multi-agent workflows with tool use and RAG
• Improved ranking correlation from 0.716 to 0.821

CorvEd

Full-stack tutoring operations platform covering intake, payments, matching, and scheduling.

• Next.js App Router
• Supabase Postgres
• Row Level Security
• PKCE OAuth

WorkWell Measure Studio

Production healthcare compliance platform with deterministic CQL evaluation, interoperability, and guarded AI assistance.

• TypeScript
• FHIR R4
• SMART Backend Services
• CQL
• MCP
• PostgreSQL

TurboQuant

PyTorch implementation and evaluation of KV-cache compression for long-context LLM inference.

• 5.2x validated cache compression
• Qwen retrieval-safe evaluation through 32K context
• PyTorch
• Hugging Face DynamicCache

RASS / CoRAG

Multi-service retrieval platform with asynchronous ingestion, hybrid search, grounded streaming answers, citations, and observability.

• Node.js
• OpenSearch
• Redis
• PostgreSQL
• Docker
• MCP and SSE

UNLP 2026 Ukrainian Document QA

Retrieval-augmented question answering pipeline combining BGE-M3 retrieval, Qwen3 reranking, and Gemma-based multiple-choice answering.

• Private score 0.8722
• Ranked 7/15
• BGE-M3
• FAISS
• Qwen3