AI governance + analyst-in-the-loop oversight demo using open-source geo/text data

Cloud-independent flood detection using Sentinel-1 SAR data via Microsoft Planetary Computer. Demonstrates before/after radar backscatter change analysis to map flood extent under all weather condi…

Mapping VSAIW (Scaled Agile) to Responsible AI controls

Simulated GitLab compliance controls pipeline for IC use case

Demonstrates adversarial attacks on transformer models using TextAttack and HuggingFace Transformers

A demonstration of how GitLab’s group-project inheritance and agile planning features support multi-organization delivery coordination under Scrum

Demonstrates GitLab CI/CD pipeline with parameterized deployments (env, version, toggles)

This project demonstrates the use of `plotnine` (Python's ggplot2 equivalent) and `seaborn.swarmplot` (Python's beeswarm plotting tool) to visualize synthetic datasets.

Policy-driven benchmarking of language models using Stanford HELM evaluations.

Demonstration of the industry shift from AI safety (bias, fairness) to AI security (threat prevention, adversarial risks)

Demonstrating Python upskilling through data cleaning, visualization, time-series forecasting, geospatial mapping, and fairness evaluation

Prototype chatbot assistant for AI governance, aligned with NIST AI RMF and Responsible AI policy support.

Demonstrating basic geospatial data analysis and visualization using GeoPandas to plot threat incidents and create simple heatmaps.

Statistical modeling and forecasting of synthetic threat incidents using Poisson regression, ARIMA, exponential smoothing, and panel data techniques. Built to demonstrate applied data science skill…

Applied statistical models, time-series analysis, spatial clustering, and real data integration (GTD, HIFLD) for simulating and analyzing threat patterns.

Full-lifecycle AI model development demo aligned with responsible AI and ODNI governance

AI bias evaluation in autonomous vehicle systems using Fairlearn. Based on prior Public Policy PhD research in AI & automation ethics.

Demo of interpretable, transparent, and steerable safety moderation of AI behavior, inspired by SafetyAnalyst (2025)

AI Safety Assurance demo project applying AMLAS, IEEE Safety Assurance, ACM Taxonomy, and Explainable AI techniques.

A demonstration dashboard showcasing AI governance maturity, public sector AI readiness vs. responsibility, fairness and bias assessments, and adversarial AI risks — aligned with NIST AI RMF, Micro…

Explainability and AI risk evaluation for generative AI outputs, using SHAP visualizations and risk analysis aligned with Responsible AI practices.

Evaluation of generative AI chatbot responses to public service queries using Microsoft Azure AI Foundry — applying AI quality, NLP, and risk/safety metrics for responsible AI oversight.