Tony Siress

I spent three decades in enterprise technology—Sun Microsystems, Veritas, Motorola—watching systems scale until they broke and learning what theory never teaches. Then another two decades in venture capital, pattern-matching across hundreds of pitches to spot which architectures would hold and which would crack under pressure.

Now I'm applying both lenses to AI system design.

My AICred assessment revealed something I'd suspected: I don't just use prompts, I engineer them. I build constraint architectures that prevent fact drift before it happens, design multi-stage chains that separate concerns cleanly, and make template selection decisions with the diagnostic precision of someone who's debugged enough production failures to know where problems actually originate.

The gap between knowing AI concepts and deploying AI systems that work at scale? I operate on the deployment side of that gap.

My next frontier is stress-testing—deliberately pushing my architectures past their limits to discover failure modes that only volume reveals. It's the same instinct that made me effective in enterprise tech: find where it breaks before your users do.

I spend much of my time at my lake house in Ajijic outside of Guadalajara, in Jalisco Mexico now, after five decades in Silicon Valley. and 6 years in Austin Texas. I am slowing exiting from active investing in VC, but not from building.

The first person makes it direct. No one else narrating your story.

Tony Siress brings five decades of enterprise technology and venture capital pattern-matching to AI system design. Three decades at Sun Microsystems, Veritas, and Motorola taught him where systems break under pressure. Two decades evaluating hundreds of startup architectures sharpened his instinct for what holds and what cracks.

His assessment revealed engineering-grade prompt architecture: constraint systems that prevent fact drift before it happens, multi-stage chains with clean separation of concerns, and diagnostic precision forged from years of production debugging. He operates on the deployment side of the gap between knowing AI concepts and building AI systems that work at scale.

Now based at his lake house in Ajijic after five decades in Silicon Valley, Tony isn't slowing down. He's stress-testing his architectures to find failure modes before users do.