Case Studies
In-depth breakdowns of software architectures, machine learning models, and system implementations.
SkyRipple: Agentic Airtraffic Simulator
Operating on 582,304 real BTS flight records, 190,711 crew pairings, and 2.41 million synthetic passenger itineraries, SkyRipple is an analytical discrete-event delay propagation prototype and agentic recovery simulator. This case study outlines the technical implementation of the heavy discrete-event engine, the LLM boundary guardrails, the multi-agent arbitration architecture, and how the simulation was compiled into a zero-latency static web demonstrator.
Agentic Systems for Synthetic Healthcare Data
A synthetic data generation system that utilizes agentic workflows to simulate realistic patient-practitioner interactions, orchestrating stateful agents to generate structured, FHIR-compliant healthcare records.
Autonomous Mapping: Neural Spatial Conflation
A two-tower Neural Network that automates the spatial linking of traffic assets to physical mounting structures, identifying LiDAR sensor anomalies and accelerating map update velocity.
AI-Driven Faculty Research Classification
Developed an agentic pipeline using LangGraph and Gemini Flash to categorize unstructured university-wide research descriptions into standardized taxonomies.
MediQuery: AI Patient Triage & Provider Routing
An AI-assisted health navigation prototype that processes natural language symptoms to triage conditions and route users to localized specialists.
Chicago Diabetes Modeling: Spatial Predictive Analytics
Quantified socio-economic determinants of adult diabetes across 77 Chicago community areas using regression and spatial clustering, deploying an interactive dashboard with AI-assisted interpretation.
Predictive Volatility: Election Forecasting Models
Evaluated national election forecasting models using custom probabilistic stress-tests, Brier score decompositions, and conditional dependencies in R and Python.