Each project explains the problem, my contribution, the technical approach, the result, and what I learned.
Independent portfolio work demonstrating how I design, test, measure, and communicate AI-enabled products.
Source-grounded publishing intelligence with research agents, ingestion and review workflows, PostgreSQL, prompt governance, execution telemetry, AI evaluation, deterministic safety, and human review.
A focused technical case study on deterministic LLM evaluation, regression protection, request tracing, tokens and cost, latency, failures, workspace budgets, dashboards, and quality gates implemented as part of the WPH platform.
A practical framework for assessing AI opportunities, data readiness, governance, risks, controls, and organizational change.
Graduate projects translated into recruiter-friendly case studies with real methods, results, limitations, and next steps.
Compared linear and tree-based models, evaluated RMSE, MAE, and R², and used residual analysis to identify production and fairness risks.
Provisioned a cloud environment, ingested NHTSA transportation data, validated transformations, and documented a repeatable workflow using ARM templates and deployment parameters.
Evidence-based summary of defect analysis, test coverage, release readiness, automation contribution, and quality leadership—the foundation behind how I approach AI reliability.
Start with the product demo, then go deeper into architecture, AI evaluation and safety, or the recruiter briefing.