AI Quality Engineer with 7+ years across compliance-regulated systems (HIPAA, ISO 27001, NIST 800-63), now focused on LLM evaluation and agentic AI quality engineering. Author of Verdict (verdict-eval on PyPI) — an open-source evaluation harness for AI agents with injection-resistant LLM-as-judge, HIPAA/NIST compliance modules, and CI/CD quality gates. Completing M.S. in Computing & Data Sciences at Boston University (GPA 3.91). Seeking remote LLM Evaluation Engineer or AI Quality Engineering roles.
LLM evaluation, agentic AI quality, and full-stack test automation
7+ years of building quality into products that matter
Applied engineering at the intersection of QA and AI
Open-source evaluation harness for AI agents with injection-resistant LLM-as-judge, HIPAA/NIST compliance modules, and CI/CD quality gates — published on PyPI as verdict-eval
Four-agent evaluation architecture (test generation, execution, judgment, reporting) using CrewAI and the Anthropic SDK, with a pluggable adapter pattern supporting any AI system — LangChain, LangGraph, raw API, or custom agents. Achieved 22/22 pass/fail agreement and 5/5 critical failure detection on a hand-labeled calibration set across all 5 test categories.
Autonomous pytest generation from requirements documents & OpenAPI specs — Postman alternative powered by Claude AI
Feed it a requirements doc (txt, md, PDF, docx) or an OpenAPI/Swagger spec. It reads the endpoints and user stories, generates a comprehensive pytest suite, runs the tests, and if any fail, sends the tracebacks to Claude for root cause analysis. Works without an API key — mock AI produces structurally valid tests for prototyping and CI pipelines.
@pytest.mark.hipaa_164_312_d decorators and generates gap reports
Autonomous 20-agent stock signal platform built on LangGraph & Claude API — real-time BUY/SELL signals with WhatsApp alerts via Twilio, deployed 24/7 on DigitalOcean
20 specialized agents run in parallel around the clock on a LangGraph pipeline — macro analysis, sentiment, breadth, and signal generation agents each handle a distinct role with structured handoffs. Integrates Polygon.io and yfinance for real-time market data, delivers BUY/SELL/HOLD signals via WhatsApp (Twilio), and runs as a systemd service on DigitalOcean — auto-restarts on crash, survives reboots, no manual intervention.
Designed a suite of tests to verify service deployment and scaling behaviors within containerized environments, ensuring reliability across distributed microservices.
Built automation templates for React Native and Flutter SDKs, enabling faster cross-platform validation across multiple client implementations at Aware, Inc.
ML modeling, statistical analysis, and responsible AI applied to real datasets
Built classification models using Random Forest and Neural Networks for anomaly and threat identification across network traffic data.
Built an LLM workflow using the Gemini API to route homeowner service requests. Designed short vs. structured prompt variants for controlled comparison; used adversarial examples to expose keyword-bias weaknesses.
Feedback from colleagues and managers on LinkedIn
Dan is an exceptional QA engineer who brings both technical depth and genuine ownership to every project. His ability to design comprehensive test strategies — from API-level automation to end-to-end biometric validation — is outstanding. He consistently raises the bar for quality on the team.
Working with Dan at Mobile Heartbeat was a great experience. He is meticulous, fast, and deeply committed to product quality. His HIPAA testing expertise and ability to catch edge cases in our VoIP communication stack saved us from multiple production issues. A true professional.
Academic foundations and continuous upskilling
Seeking remote LLM Evaluation Engineer or AI Quality Engineering roles