Patented AI Innovator in Regulated Environments
Production Systems
- ▸Spark AI Platform: 50+ purpose-built enterprise tools across learning, productivity, sales, HR, and creative workflows. Saving millions in time, effort, and cost.
- ▸Enterprise RAG: Built and deployed in 1 month. Vendor quote was 6 months and over $250,000.
- ▸Multi-LLM orchestration: Routes tasks to the best model for the job and ensures agents have access to the right tools, data, and context. OpenAI, Anthropic, Google, and Perplexity unified under one layer.
- ▸Rapid prototyping: Critical projects delivered in hours, not months. Secured a $1M+ grant by building production web pages in under 4 hours.
Governance & Compliance
Built in, not bolted on. Custom tools with deep integrations into governance systems, ensuring adherence to the most current policies and regulations.
- ▸AI governance frameworks for a global automotive and aerospace standards organization
- ▸Data privacy architecture for sensitive engineering data
- ▸Patent holder (#11520470) for AI-powered UX systems
- ▸Co-authoring organizational AI and agentic governance policies and frameworks
Ask My AI
Curious about my background? Ask anything. Type or use your voice.
Professional Summary
Agentic AI Systems Leader with 20+ years strategizing, leading, and delivering at the intersection of people and technology. Currently directing enterprise AI strategy and orchestrating 50+ production AI systems at SAE International, a global automotive and aerospace standards organization.
- ▸Directing AI agent teams, not just building them. Evaluating output, setting strategy, and orchestrating multi-agent workflows that run in production daily
- ▸Architected enterprise RAG platforms, deployed customizable AI assistants, orchestrated agentic AI systems, and shipped Chrome extensions to the Web Store
- ▸Bridges the gap between technical AI capability and business strategy, from regulated industries to independent product launches
- ▸Patented strategist delivering innovation in both regulated and unregulated environments
Real-Time Agent Build
A simple demonstration of multi-agent collaboration. Three AI agents coordinate to build a landing page using real API calls and real-time orchestration. This is just a glimpse of what becomes possible with more complex agentic systems, from scheduled automation and regulatory compliance monitoring to full enterprise workflows, all built for safety and reliability in regulated environments.
Preview will appear here...
Agent Roundtable
Four AI agents. Real API calls. No scripted responses. Every answer is generated live by different LLMs debating your chosen topic. Nothing here is pre-written. This is multi-agent coordination for strategic thinking, not just task execution.
Select a Topic
Multi-Agent Orchestration in Production
A visual representation of the multi-agent architecture I design and deploy in production. Specialized agents coordinate in real time, each with defined roles, tool access, and handoff logic. This is the pattern behind the systems I build. Explore the diagram to see how it fits together.
The Agentic AI Spectrum
Most organizations confuse LLM tools with Agentic AI. The boundaries matter.
LLM Tools
- • Conversational interfaces
- • Knowledge retrieval
- • Content generation
Click to expand
Copilots
- • Contextual assistance
- • Human-in-loop
- • Augmentation focus
Click to expand
Workflows
- • Multi-step automation
- • Deterministic paths
- • Process optimization
Click to expand
Autonomous Agents
- • Goal-driven behavior
- • Self-correcting
- • Minimal oversight
Click to expand
Technical Foundation
Hands-on experience across all modern AI stacks
LLMs & Embeddings
- • Production deployment across OpenAI, Anthropic, and Google model families
- • Custom embedding models for domain-specific search
- • Prompt engineering at enterprise scale
RAG & Vector DBs
- • Built enterprise RAG system (Google Vertex AI)
- • Advanced vector search for enterprise knowledge bases
- • Hybrid search (semantic + keyword)
Agent Orchestration
- • Production-grade multi-agent orchestration at enterprise scale
- • Deterministic vs probabilistic agent design
- • Real-time agent coordination and handoffs
The Journey
Each chapter built the foundations for the next - from creativity to scale, regulation to AI orchestration.
- ▸Flash Animation Pioneer (Late 90s): Microsoft/Advanced Learning, Galway Ireland - Windows Server 2000 training content that reached global audiences
- ▸NYC Government (2001-2013): 9/11 emergency response, systems serving millions of New Yorkers, NYC Excellence in Technology award - learned to build at scale under pressure
- ▸Enterprise UX Leadership (2013-2017): DDI - led mobile strategy and enterprise CMS architecture, bridging user needs and business constraints
- ▸SAE International (2017-Present): From UX lead to Agentic AI Systems Leader. 50+ production AI systems, patent holder, architect of AI governance frameworks for a global standards organization
Career Highlights
SAE International (2017-Present) - Principal UX Designer & AI - 50+ production AI systems, patent holder, enterprise AI at 98% less cost, AI governance architect
DDI (2013-2017) - UX/UI Design Lead - Mobile strategy, enterprise CMS
NYC ACS (2005-2013) - Webmaster & Sr. Manager - NYC Excellence in Technology award, systems serving millions
NYC DOITT (2001-2005) - Web Design Consultant - 9/11 emergency response, thanked twice by Mayor
Microsoft/Advanced Learning (1999-2000) - Flash Animator - Windows Server 2000, Galway Ireland
Key Projects
Spark AI Platform
ProductionEnterprise AI hub spanning learning, productivity, sales, HR, and creative workflows. Multi-LLM orchestration underneath routes tasks to OpenAI, Anthropic, Google, and Perplexity based on capability and cost. 50+ purpose-built tools with built-in governance.
Enterprise RAG System
ProductionBuilt in 1 month for ~$4K versus external vendor quote of 6 months at $250K - 60x cost savings, 6x faster delivery. Google Vertex AI with custom embeddings and hybrid search for enterprise knowledge retrieval.
Multi-Agent Orchestration
ActivePersonal AI system with persistent memory, autonomous sub-agent spawning, heartbeat monitoring for proactive assistance, and voice synthesis for natural interaction. Agents coordinate, delegate, and self-heal with minimal human oversight.
VIP Chrome Extension
PublishedVoice and vision AI-powered Chrome extension using vision AI for screen analysis and natural language interaction. Published in Chrome Web Store - real-world deployment with user authentication and privacy controls.
AI Governance Framework
CompleteComprehensive AI governance documentation for SAE International with EU AI Act awareness, data privacy architecture for sensitive engineering data, and risk assessment frameworks for regulated deployments.
Patent: UI for Digital Standards
GrantedPatent #11520470. Novel AI-powered UX for interacting with digital engineering standards - making complex technical documentation more accessible and actionable through intelligent interface design.
Impact Metrics
Get in Touch
Interested in working together? Drop me a message and I will get back to you within 24 hours.
Or email me directly at john@agentjmb.com