Principal Enterprise Architect AI/ML and independent researcher in AI governance and safety.
Currently contributing to apache-airflow-providers-common-ai.
I design enterprise AI/ML platforms and research what breaks when LLM systems operate in regulated settings. 18 years in IT, with the last several years focused on enterprise architecture for AI/ML platforms and multi-agent AI governance. Currently building AI/ML reference architecture across Databricks, Azure AI Foundry, and Snowflake at enterprise scale. Independently researching multi-agent LLM safety, emergent misinformation, and governance frameworks for regulated industries.
AI Architecture Enterprise Patterns 18 enterprise AI architecture patterns with 228 interactive data flow visualizations. Covers unified AI gateway, RAG for regulated data, multi-agent safety gates, compliance-aware routing, governance-as-architecture, contamination-resistant pipelines, FinOps, security, observability, and more. Zero install required. Explore live demos.
GAIF Governance Observatory Open-source governance toolkit for the Governed AI Architecture Framework (GAIF). Six interactive CLI tools covering governance maturity assessment, risk scoring, compliance gap analysis, and architecture fitness evaluation. 25 passing unit tests.
Contamination Percolation Research codebase for measuring how errors propagate between AI models in multi-agent pipelines. Introduces the T1PR (Tier-1 Percolation Rate) diagnostic and demonstrates the gap inversion effect across DBRX, Claude, Llama, and Gemini model families.
My independent research focuses on what breaks when LLM systems operate at scale. All work is solo-authored.
| Paper | Status | Venue |
|---|---|---|
| EMG (Emergent Misinformation Genesis) | Preprinted | Targeting NeurIPS 2026 D&B |
| PHI-GUARD (Compliance-Aware LLM Routing) | Under review | IEEE JBHI |
| ContamPerc (Contamination Percolation) | Under review | IEEE Access |
| MedMI-Bench (Clinical MCQ Benchmark) | Under review | JMIR AI |
| GEG (Governance Effectiveness Gap) | Under review | BCM MDIM |
| TEMPORAL-MED (Temporal Consistency) | Targeting | JMIR AI |
GAIF Framework: GAIF-4 v1.5 defines four quantitative metrics for AI governance health: EMR (Emergent Misinformation Rate), T1PR (Tier-1 Percolation Rate), CFR (Compliance Failure Rate), GDR (Governance Decay Rate).
- EMG: Zenodo
- PHI-GUARD: TechRxiv
- GAIF: Zenodo, SSRN
- GAIF-4 v1.5: Zenodo
- IEEE Big Data 2026 Program Committee (Industry and Government Track)
- IEEE Access Reviewer
- NIST Public Comments: CAISI Framework, NCCoE AI Guidelines
- MCP Enterprise Interest Group: Healthcare and Compliance Use Cases Champion
- MS in progress, Colorado Technical University
- Newsletter: AI Architecture and Governance
- ORCID: 0009-0005-5107-4485
- Email: [email protected]