By Julie Galante
Cesar R. Hernandez MC/MPA 2026 is building a democratic “civic operating system” so smart cities of the future are equitable, accountable, and human-centered.
—
When Cesar R. Hernandez MC/MPA 2026 talks about his work in artificial intelligence, democracy, and smart cities, he doesn’t begin with technology. He starts with his family.
Hernandez’s family are K’iche’, an Indigenous Mayan people from the central highlands of Guatemala who resisted Spanish colonization. “If you ask them in Guatemala if they’re still fighting, they’ll say, ‘Yes, this is our land,’” he says. That generations-long struggle against colonization shaped his family—and eventually, his path to Harvard Kennedy School.
His great-grandfather fought in secret wars against the United Fruit Company. His grandmother organized pro-democracy campaigns in Guatemala City in the 1950s; when her campaign friends began disappearing and turning up in landfills, she fled to the United States.
“These weren’t just stories,” Hernandez recalls. “As I got older, I realized, ‘This is actual history.’ My family has talked about struggle, survival, and mass movement for generations, and now here I am, born and raised in this country. I didn’t have to worry about survival or colonization or the same kind of political oppression. I’m in a place where I can create.”
That awareness drives his work.
From community organizer to civic technologist
Hernandez’s journey into smart cities and agentic AI didn’t begin in a lab, but in community centers and U.S. congressional offices.
After college, he worked as a community organizer in Florida, helping Indigenous Mayan families navigate schools and social services. He found himself lobbying U.S. Congress members on their behalf.
“The reality is, they don’t have a voice,” he says. “They don’t have voting power or political power.” Lawmakers praised his work, took photos, and moved on. “I asked myself, ‘How can I be part of this process so I’m not just getting thank-yous?’”
That question pulled him into government, where he became a legislative aide and saw how policy is made—and where it falls short. From there, he shifted into political consulting, hoping to help elect better leaders only to discover how unpredictable candidates can be once they enter office.
The lesson, he says, wasn’t that politics didn’t matter. It’s that politics alone isn’t enough.
“You need what I call ‘navigators’ or ‘orchestrators’ of society,” he says—people who understand communities, government, markets, and technology and can align them toward shared goals.
That conviction led him into transportation and technology to connect marginalized communities to jobs and opportunity. In Tampa Bay, Florida, he helped design a partnership between a rideshare company, a transit agency, and Tesla Motors. The collaboration increased transit ridership by 80 percent and linked transit-poor neighborhoods to major job centers.
Hernandez says it was a turning point.
“If we can use technology through public-private partnerships, government, and the community—especially where the bureaucratic model is slow—we can design models that give people what they need,” he says. “That’s where my obsession with tech and equity came from.”
He went on to work on policy frameworks for autonomous vehicle pilots, rideshare models, Hyperloop systems, and advanced air mobility, when a bigger question began to form: What if these systems could be orchestrated together?
Building Omni | X and IRIS, a nervous system for cities
In 2017, Hernandez launched a startup concept called Omni Mobility, an ambitious effort to connect autonomous vehicles, aerial vehicles, and nautical drones into a unified network. The vision was early, but the technology wasn’t ready. The company failed.
He went back to government relations and public affairs. Then generative AI and agentic systems started gaining traction.
“I realized that was the solution about a year ago, with advances in AI and agentic systems,” he says. “Smart cities are fragmented. You have one dashboard for traffic lights, another for utilities, another for infrastructure. These interdependent datasets don’t speak to each other; it creates fragility.”
Hernandez’s company, Omni | X—now based at the Harvard Innovation Labs (iLabs) and incubating at the MIT Media Lab’s AI Venture Studio—is working to change that. Omni | X is building a unified AI architecture that connects drones, robots, and critical infrastructure systems in what Hernandez calls an “agentic layer” for cities. While initially focused on civic infrastructure, he says Omni | X is also being adapted for counter unmanned aircraft systems (UAS) coordination to help frontline operators track and respond to autonomous aerial threats using a single, unified AI system.
At its core is IRIS, a “mega-agent” he developed at the MIT Media Lab. It’s an engine that can ingest and navigate multiple datasets and systems with one human-supervised layer.
“The beauty of an agent is that I can give it my email, my HR documents, my permitting documents, my traffic-light information, procurement data, and even access to drones—all in one agent. That’s what Omni | X is,” he explains.
Transportation departments are asking Hernandez to use IRIS to solve concrete problems like centralizing and managing permit-issuing workflows. On the other side of the world, he is in discussions with the Kingdom of Bhutan about using IRIS to help design Gelephu Mindfulness City and train an agent on Buddhist philosophy and culture so spiritual values are embedded in the new city’s operating system.
“Agentic AI is about a society of agents,” Hernandez explains. “If I build an agent for City X and an agent for City Y, and both are on the Gulf of Mexico and a hurricane is coming, those agents can instantly communicate with each other. They’ll say, ‘This is what we’re seeing in real time. This is what’s happening based on the drones we’ve deployed.’”
His long-term vision is a “civic OS”—an operating system for civilization that defines how AI lives in cities, interacts with people, and is governed.
Guardrails for the civic AI era
Hernandez fervently believes in human-in-the-loop systems. “There must be a human composer to the orchestra of all these datasets and tools,” he explains. Without that, he worries, AI-driven infrastructure could become an unaccountable force.
“If we build systems without human-in-the-loop oversight to manage drone systems—and those drones are managing communities that look like mine—who’s in charge when a drone misfires? We have a huge accountability problem if the answer is, ‘It’s not me, it’s the code.’”
He points to the rise of what he calls “authoritarian AI”—turnkey smart-city deployments by countries that bundle infrastructure upgrades with data access.
“We’re in an unseen battle over how artificial intelligence will be used in what I call the civic AI era, where AI must live in some kind of harmony with humanity.”
The moral imperative, he says, is to build “democratic AI”—systems with auditability, safeguards, and clear lines of responsibility.
“Success,” he says, “means building an outstanding product that can orchestrate society, but one that lives in harmony with us and builds a society that is equitable, has clear guidelines and triggers, and co-exists with AI in a healthy way.”
“We’re in an unseen battle over how artificial intelligence will be used in what I call the civic AI era, where AI must live in some kind of harmony with humanity.”
Why HKS and why now
Hernandez describes his decision to come to HKS as improbable and inevitable.
He’d “always wanted to go to an institution where knowledge collapses time,” he says. “You’re surrounded by people who are knowledgeable in their arenas of expertise and willing to lend that expertise to help push your vision forward. I feel I’ve gained 10 to 20 years in the past six months just being in Cambridge.”
Initially, he planned to attend a different program to “make himself competitive” for Harvard—until a friend challenged him to apply. He did, was admitted to the MC/MPA Program, and was named an Equity Fellow at the Center for Public Leadership (CPL).
For him, Omni | X is not just a startup.
“I’m not trying to aggrandize my career or find a job,” he says. “I’m looking to gain knowledge, wisdom, and understanding so I can return to my community and help build a more equitable future. That’s a mandate for me.”
He is candid about how instrumental HKS, CPL, and the Harvard Innovation Labs have been in turning that mandate into a real venture.
“This would have been impossible without HKS, without the Innovation Lab, without CPL,” he says. “I meet with CPL staff at least twice a semester, and they have a checklist: ‘How’s it going? What do you need? Who should we connect you with?’ They’re very methodical in making sure that when you come here, you’re successful.”
Through HKS, he cross-registered at MIT and joined the AI Venture Studio at the MIT Media Lab—a competitive program where he became a finalist as a non-technical founder. There, he built the first version of IRIS and recruited MIT developers and teaching assistants to join him. Today, Omni | X is recognized as a “minted” MIT Media Lab company and works with a development firm that has built tools for the U.S. intelligence community.
Now, with letters of intent from cities and departments of transportation and pilot programs in development, he’s focused on turning his technical and policy groundwork into impact.
“When I talk to developers and say ‘MIT Media Lab,’ they say, ‘Oh, you’re legit,’” he notes. “When I say ‘CPL’ to a government official, they say, ‘Oh, you’re vetted.’ Those layers of credibility and validation are extremely helpful, especially when you’re building technology that interacts directly with human beings and can create friction.”
Looking ahead (100 years)
When asked what success looks like to him, Hernandez is still working on the answer.
Initially, he imagined the classic Silicon Valley path: build, scale, exit, move on. But his time at HKS and in Cambridge complicated things. Learning about century-old Japanese companies—Japan has more than 30,000 firms over 100 years old—led him to a new question: What does it look like to build a company that lasts 100 years?
He knows that one day he wants to run for office—and to be self-funded when he does. For now, he is channeling his family’s history of struggle into building the civic AI infrastructure he believes the future demands.
“So that’s what I’m doing,” he says. “It could all blow up in my face, but I have no plan B.”
Images courtesy of Cesar R. Hernandez MC/MPA 2026
