00
Projects
Projects across recommendation systems, autonomous learning, LLM evaluation, predictive analytics, and healthcare AI.
01
Case Studies
01
Explainable booking intelligence for matching artists with venues and local demand.
Deployed recommendation platform using live data across 336 artists, 52 venues, and 428 artist-event relationships.
Recommendation Systems / Explainable Scoring / Data Pipelines
02
Physics-based simulation and imitation learning for robust racing-line control.
Physics-based simulator and imitation-learning pipeline with behavior cloning, DAgger-style correction data, and zero-shot transfer testing.
Imitation Learning / DAgger / Robustness Testing
03
Source-grounded equity research with explicit validation and deterministic fallbacks.
Structured equity-research workflow with source-quality checks, hallucination analysis, unsupported-claim detection, and deterministic fallback paths.
LLM Evaluation / Hallucination Analysis / Source Grounding
04
AI-powered residential real estate prediction system for sale price and sale timing analysis.
Processed and engineered 200+ predictive features, evaluated observed outcomes, and translated recurring model failures into pricing insights.
Predictive Modeling / Real Estate Analytics / Sale Timing Analysis
05
Machine-learning fraud decision support with a companion real-time risk monitoring layer.
Combined machine-learning fraud scoring and investigator routing with an event-driven companion layer for provider-level fraud and cyber risk alerts.
Fraud Detection / Decision Support / Responsible AI
06
Risk-aware portfolio analytics workflow for benchmarking, simulation, and allocation analysis.
Portfolio analytics workflow covering Sharpe ratio, beta, drawdown, benchmark comparison, and Monte Carlo simulation.
Portfolio Optimization / Sharpe Ratio / Risk-Adjusted Returns
07
Classification workflow for predicting perfect guest rating outcomes from listing features.
~0.815 validation ROC-AUC across structured and text-based listing features.
Classification / Feature Engineering / Model Evaluation
08
Repeatable tournament automation framework for evaluating chess bot behavior and results.
Python tournament framework with move validation, PGN logging, JSON summaries, and repeatable bot evaluation.
Simulation / Performance Tracking