Research
Research
I work at the intersection of applied machine learning and higher-education strategy — natural language processing, knowledge graphs, reinforcement learning, and optimization, applied to the problems institutions actually face: aligning curricula to the labor market, reading signal out of unstructured disclosure data, and allocating scarce resources like financial aid more effectively.
The work below spans peer-reviewed-style empirical studies, a critical review of frontier research, and applied simulations. Several of these directly inform Allocate, the higher-education optimization platform I'm building through PraxisIQ.
May 1, 2026
Optimizing Institutional Financial Aid Discount Rates with Contextual Multi-Armed Bandits
A contextual multi-armed bandit that learns the revenue-optimal institutional aid discount per student segment, demonstrating that the yield-maximizing discount is not the revenue-maximizing one.
Peer ReviewApril 5, 2026
Peer Review — Multi-Agent Digital Twins for Strategic Decision-Making Using Active Inference
A critical review of Mancinelli et al. (2026) on multi-agent digital twins using active inference, weighing its contribution to agent coordination against computational cost.
Knowledge GraphsMarch 10, 2026
Modeling the Evolution of AI Language in Corporate Disclosures: A Workday Knowledge Graph
A single-firm longitudinal knowledge graph that tracks the evolution of AI-related language in Workday's disclosures and tests whether that language predicts returns.
NLPDecember 15, 2025
Measuring Curriculum–Labor Market Alignment in Vocational Education with NLP
A semantic information-retrieval pipeline that quantifies how closely vocational program curricula align with employer demand expressed in online job postings.