Translating Human Behavior Into Data-Driven Impact
I bridge behavioral science, machine learning, and organizational strategy to build systems that understand how people learn — and change how institutions serve them.
🏢
Professional Experience
2022 - Present
Machine Learning Data Scientist — Lexia Learning
Transforming how an educational technology organization uses data — from restructuring legacy systems to enable real-time monitoring of student learning and engagement, to building enterprise-scale AI pipelines that surface actionable insights for students, educators, and leadership. Serving as the connective tissue between software engineering, AI, and curriculum teams to ensure Automatic Speech Recognition enabled learning experiences meet complex pedagogical requirements.
2015 - 2021
PhD Researcher, Human-Computer Interaction — Carnegie Mellon
Addressed a fundamental gap in educational technology: systems that teach but can’t sense why learners struggle. Developed psychometric and data mining methods that translate motivational constructs — self-regulation, diligence, engagement — into real-time computational measures for adaptive learning systems.
2013 - 2015
Graduate Researcher, Protolab — Carnegie Mellon
Investigated principles of human creativity within complex socio-technical systems — how people generate, share, and build on ideas when mediated by technology. Designed and ran human studies while building the full-stack web platforms (Node.js, JavaScript) that made them possible. Published findings at CHI and CSCW.
2007 - 2011
Software/Systems Engineer — Lockheed Martin
Led geographically distributed teams building mobile architectures and augmented reality systems for defense applications. Coordinated cross-functional teams across hardware, software, and design — resulting in a $1.5M prototype award.
🎙
Speech Recognition Evaluation System
Production
Formalized the expert judgment behind educator speech proficiency assessments for English language learners into ML requirements through cognitive task analysis. Built a full evaluation pipeline including a 30-person labeling operation, benchmark datasets, and age-bias detection — achieving 20%+ accuracy improvement and 40% reduction in word error rate.
PythonReactTypeScriptDockerMongoDBFull-Stack Web
📈
Data-Driven Curriculum Revision
Production
Integrated educational data mining methods — learning curve analysis, knowledge tracing, and formative learning visualization — into curriculum team workflows to drive more efficient, evidence-based content revision. Reduced review scope by over 70% through empirical learning efficacy analysis and identified a significant high-need student subpopulation for targeted support.
PythonLearning CurvesKnowledge Tracing
🛠
Data Engineering for Learning Analytics
Production
Built the case for a data quality and instrumentation initiative that strengthened the foundation for all downstream modeling and analytics. Secured organizational commitment to prioritize infrastructure investment — achieving 30%+ reduction in data retention costs and enabling new capabilities in knowledge tracing and early learner intervention.
PythonRSQLPandasExploratory Data AnalysisLuigiSQLAlchemy
📝
LLM-Powered Content Generation
Production
Partnered with curriculum authors and content experts across multiple projects to decompose educational content authoring tasks — analyzing the cognitive and pedagogical structure behind each task to inform prompt design and context engineering for LLM-based generation. Applied computational linguistics theory to design scalable benchmarks for evaluating generation quality across vocabulary complexity, sentence structure, and discourse coherence.
LLMsContext EngineeringAgentic AI SystemsNLPPower Automate
☁
Cloud ML Pipeline Automation
Production
Designed flexible AWS cloud infrastructure for deploying machine learning and analytics data pipelines at scale. Built a microservice-based framework that supports arbitrary analysis and reporting workloads — from automated KPI monitoring to model evaluation — with FERPA-compliant data governance built in from the ground up.
AWS ECSLambdaStep FunctionsS3RedshiftGlueDockerCloudFormation
🧠
Measuring Motivation from Log Data
Research
Developed methods that enable learning systems to detect student motivation in real time — not through surveys or self-report, but from natural behavioral patterns in log data. Grounded in theories of self-control and expectancy-value theory, the approach leverages the contexts in which behaviors occur as discriminative indicators of latent motivational states, extending Item Response Theory for automated measurement.
PsychometricsSimulationPythonEDM
🤝
Human-AI Collaboration for Data Science
Research
Built tools enabling non-data-scientists to leverage advanced ML through human-AI collaboration. Conducted think-aloud studies contrasting novice and expert analytical processes, then designed interfaces that bridged the gap.
AutoMLUser ResearchWeb Development
🎓
Personalized Learning Squared
Research
Used quantitative and qualitative research methods to support the development of an online application that supports educators in identifying and supporting the motivational and cognitive needs of students through analysis and monitoring of students’ usage of learning applications.
Learning AnalyticsDashboard DesignUX ResearchFocus Groups
🧠
Behavioral Science & Measurement
- Motivation & Self-Regulation Theory
- Psychometrics & Item Response Theory
- Cognitive Task Analysis
- Latent Variable Modeling
- Human Studies
📊
Machine Learning & Data Science
- Supervised / Unsupervised / Semi-supervised ML
- Knowledge Tracing (Deep/Logistic)
- NLP & Speech Recognition
- Deep Learning (PyTorch)
- LLM Agentic System Design
🎯
Technical Leadership & Influence
- Cross-Functional Team Leadership
- Organizational Change Advocacy
- Stakeholder Management
- Business Case Development
- Requirements Engineering
💻
Software Engineering & Infrastructure
- Python / R / TypeScript
- Full-Stack Web (React, FastAPI, PostgreSQL)
- Data Instrumentation & Pipeline Design
- AWS Cloud & DevOps
- Docker & CI/CD
Let's Build Something That Matters
I'm interested in leadership roles where behavioral science and data strategy intersect — building teams and systems that understand people, not just patterns.