Executive Profile
I don't launch initiatives.I institutionalize capability.
Embodied AI strategy · Human-Machine Trust Architect
Twenty years inside one of the world's most complex technology enterprises — putting technology into the places people work, and building the operating models and governance that make it stick. The technology moved from automation to machine learning to AI on every screen. Now it's starting to move through the workplace itself: robots with enough AI to self-correct, adapt and work from little instruction. That's the work I'm on today, and it takes the same discipline.
140,000+ is the full population served by the platforms and programs I led, across every region — about 85,000 full-time employees plus contractors and extended workforce, not headcount I managed. My direct teams were small; my program portfolios ran up to $5M. Seven functions is the span of that work: IT, Information Security, HR, Facilities, Real Estate, Legal and Communications — owning budget, vendors, governance and the decision to ship.
In my words
I've worked through every stage of how enterprise technology learned to think: first automation, then machine learning, then the AI that now lives on every screen. The next stage leaves the screen. It moves through buildings, works alongside people and acts in physical space. The hard questions it raises are about trust, not models: who is accountable when a machine acts, who approves it, and whether the people beside it want it there.
I've spent twenty years inside one Fortune 100 technology company, at the point where technology meets the places people work. Early on I owned the benefits side of HR IT, leading the full technology stack — platforms, integrations and information security — behind a connected network of employee fitness, health and childcare centers across 4+ countries, plus partner benefits from dental care in Mexico to backup childcare and mental health support. I knew HR's business as well as the HR team did. Service and technology had to feel like one experience, or people wouldn't use them. The centers and the benefits are still in place and in use today. The lesson stuck: understand the business before you touch the technology.
Next I moved into the IT program that shows customers how the company runs its own technology. As an IT Customer Success Manager aligned to a West-region enterprise sales team, I turned how our IT actually operates into practices customers could adopt and answers sellers could bring into a deal — moving deals faster and growing wallet share, with 10%+ revenue impact. I then designed and launched our IT customer-experience center, a live showcase on our production platforms, and still lead the customer delivery around it. Along the way I helped shape our operational AI strategy for IT.
Today my focus is embodied AI. Robots are good at doing the same thing over and over; the question is what happens when something changes. Embodied AI is robots with the AI to self-correct, adapt and work on their own from little instruction. And people will work with them the way they work with each other: talking, pointing, expecting to be understood. The intelligence has to allow for that, or the human element gets lost. Preparing an enterprise for that means the operating model, the safety and governance controls, and the human side of sharing a workplace with machines that decide. It's still in the lab, and I'll share outcomes when there are outcomes to share.
My edge isn't the robot. It's two decades of building technology people trust and use, then letting it disappear into the background so the experience is what they notice — and building it to run after I hand it off. If a program needs me in the room to keep running, I haven't finished it.
For introductions
Third person, ready to use in a program, panel or introduction.
Esteban Frias is an embodied AI strategist and Human-Machine Trust Architect with twenty years inside a Fortune 100 technology company. He has led the full technology stack behind employee benefits, from fitness, health and childcare centers across four-plus countries to dental, backup childcare and mental health programs, worked as an IT customer success manager alongside enterprise sales, and designed and launched a live executive showcase of IT operations. Today he is preparing the enterprise for embodied AI: robots that self-correct, adapt and work with little instruction, while keeping the human element so people can work with them the way they work with each other. A graduate of the Latino Board Leadership Academy, he speaks in English and Spanish on AI operating models, human-machine trust and Latino leadership in tech.
How I work
Principle
Understand the business before the technology
The workplace centers worked because I learned how benefits, HR and the people using them actually operate before choosing anything.
Principle
Design for the people who'll use it
Adoption is decided by the person next to the system, not the sponsor who bought it. With embodied AI, that person shares the room with it and will treat it like a coworker. The intelligence has to be built for that.
Principle
Build to hand off
If a program needs me in the room to keep running, I haven't finished it. The centers and the benefits are still in place and in use.
Principle
State outcomes only when they're real
Work in flight is labeled in flight. Figures on this site are as reported, with their scope defined.
Credentials & board
- Latino Board Leadership Academy — Hispanic Foundation of Silicon Valley, 2016. Open to board and advisory roles where technology, workforce and trust intersect.
- Management Information Systems — Menlo College, 2011–2013.
- Certified Scrum Master — Scrum Alliance, 2015.
- Featured speaker at the company's global customer conferences, North America and Latin America, in English and Spanish. Talks →