Projects
A collection of my work in software, AI safety, and community-building.
AI Failure Modes Demo ↗
Rule-based loan approval demo showing proxy bias, overconfidence-style outputs, and brittleness from hard thresholds.
Python
AI Safety
Governance
Dataset Bias Checker ↗
Lightweight CSV audit that reports representation and acceptance rates by group and flags large outcome gaps.
Python
Fairness
Auditing
AI Policy Tradeoff Simulator ↗
Toy governance simulator that scores policy choices across risk, accountability, innovation, and compliance cost (0–100).
Python
Policy
Governance
Toy Model Failure Demo ↗
Toy prediction model that outputs PASS/FAIL + "confidence" and explains limitations, omitted variables, and misuse risk.
Python
Interpretability
AI Risk