About

Who I Am

I've spent 20+ years identifying massive inefficiencies and building solutions before markets know they need them. Currently pioneering MCP (Model Context Protocol) implementations that let anyone interact with blockchain through natural conversation—eliminating the technical barriers that keep 99% of people out.

My edge: I see the intersection others miss. When I merged AI with blockchain at Andromeda, it 10x'd our engineering velocity, earning me guest lecture invitations at top universities and a keynote at the Web3 Summit in Amsterdam.

Teaching & Speaking

Guest lecturer at Harvard, MIT, Princeton, Cornell Tech, NYU, and Oxford. Keynote speaker at Web3 Summit Amsterdam. Topics covered:

• Evolution of language models: RNNs → Transformer architecture
• Modern architectures: dense vs. sparse MoE and applications
• Training and fine-tuning (cloud and local)
• Advanced techniques: RAG, Graph RAG, and AI agents
• State-of-the-art models from OpenAI, Anthropic, Google
• Integration of AI into developer tools and creative workflows
• Emerging trends: AI safety to neuromorphic chips

Track Record of Being Early

• 2002 - Built digital ticketing systems before Eventbrite existed
• 2004 - Pioneered multi-venue festivals (managed 30,000-person events)
• 2020 - Integrated NFTs into gaming before "Web3" was mainstream
• 2025 - First to build an MCP server that doesn't just query blockchain, but:
  • Automatically generates full business plans
  • Creates app/blockchain architecture plans
  • Generates wallets and resources needed
  • Deploys smart contract stacks
  • Generates the front-end, ready to connect

"Sometimes being too early is the same as being wrong, but when the market catches up, the impact is transformative."

This Blog's Journey

I first discovered MCP servers in 2024, but it wasn't until January and February 2025 that I really dove deep. In June 2025, I left the company I was with and started producing my own work—including the first round of MCP servers, notably the Osmosis server.

From July through October 2025, I built 17 blockchain MCP servers entirely by hand. Each one took 12+ hours of deep work: diving into SDK documentation, figuring out network configurations, implementing wallet operations, testing transaction flows, debugging edge cases.

This wasn't efficient. But it was deliberate.

What This Blog Documents

These 26 posts chronicle the complete evolution from manual development at scale to an automated factory system. It's not a tutorial series—it's a technical memoir. Real dates, real code, real mistakes, real breakthroughs.

• Part 1: The Chaos Era - Monolithic nightmares, TypeScript migrations, git submodule disasters
• Part 2: Discovering Patterns - Learning through repetition: logging, error handling, naming standards
• Part 3: The Automation Breakthrough - Building the MCP Factory that generates servers in 8 seconds
• Part 4: Advanced Topics - Multi-AI orchestration, PostgreSQL memory systems, crisis stories

The Numbers

The AI Collaboration

This project—and this blog—were built with AI assistance. Not just one AI, but multiple models working in parallel:

• Claude for architecture, debugging, and narrative
• GPT-5 for code review and refinement
• Gemini for systematic testing and documentation

The multi-AI orchestration approach is documented in detail in Post #15: The AI Dream Team. I'm transparent about this because AI-augmented development is part of the story, not something to hide.

Key Achievements

• MBPS v2.1 Standard - Universal naming convention for blockchain tools
• MCP Factory - Meta-automation system that generates production-ready servers
• 17+ Blockchain Integrations - Ethereum, Bitcoin, Solana, Cosmos, Osmosis, and more
• Multi-AI Workflow - Parallel AI orchestration for debugging and development

Why "Operational Semantics"?

Because the messy middle matters. Most technical blogs show you the polished final result. This one shows you the git submodule disasters, the ESM/CommonJS hell, the router DDoS blacklists, the BigInt testing nightmares—the delta between what code should do and what it actually does.

Operational semantics investigates how systems execute in practice, not just in theory. These 26 posts document that understanding: experience-driven innovation through doing the work wrong first, learning from it, then building the right abstraction.

Connect