Mat Velloso is joining Revelo as a strategic advisor.
Mat spent the last decade building developer platforms at three of the world's largest AI companies. Most recently, he was VP of Product for the Developer Platform at Meta's Superintelligence Labs. Before that, he led Google DeepMind's AI developer products, including the Gemini API, Google AI Studio, and Gemma. And he spent four years as Technical Advisor to Microsoft CEO Satya Nadella, where he was among the first few people at the company to interact with what would become ChatGPT.
We're excited to have him. Here's why he's excited about us.
The Flywheel Between AI and Engineering Productivity
There's a cycle forming in software right now that we think will define the next decade.
AI models are getting better at writing code. That's making engineers more productive. But here's what most people miss about what happens next: those more productive engineers don't just do the same work faster. They push into new territory. They build things that weren't possible a year ago. They explore new architectures, new patterns, new use cases.
And those new frontiers? They're exactly the use cases the models haven't been trained on yet. So the models need to evolve. They need expert training data from real engineers who understand the new territory. Which makes the AI better. Which makes engineers more productive. And the cycle starts again.
It compounds. Every turn of this flywheel raises the ceiling on what's possible and increases demand for both better AI and better engineers.
Why engineers become more valuable, not less
This is the part that runs counter to the popular narrative. When AI makes it cheaper and faster to build software, companies don't hire fewer engineers. They get more ambitious. They greenlight projects that were too expensive before. They expand into new products, new markets, new technical bets. Demand goes up.
And the engineers who thrive in that world aren't the ones doing the same work they did five years ago. They're the ones who know how to work with AI tools to push productivity to a level that wasn't reachable before. The premium on that skill is growing fast. LinkedIn just priced senior engineer AI trainers at $150 an hour, their highest rate.
Why frontier AI labs need real engineers, not crowd workers
On the other side of the flywheel, the models themselves need to keep up. As engineers push into new territory, the training data that got the models this far isn't enough. The models need feedback from people who actually write, review, and ship production code. People who can evaluate whether AI-generated code is correct, efficient, secure, and ready for production.
That's a fundamentally different skill than what generalist data labeling provides. It requires senior software engineers with real domain depth.
Why Revelo Is at the Center of The AI Code SuperCycle
Revelo sits at the intersection of both sides of this flywheel.
We have 400,000+ vetted senior software engineers across 50+ countries. We build the data infrastructure for AI code generation: expert human evaluation, code training environments, and model evaluation tooling. And through the same network, we provide US companies with elite nearshore engineers for their teams.
We power both sides. The engineers who push boundaries, and the expert data the models need to keep up.
Mat has watched this dynamic play out from inside the companies building the most advanced AI systems in the world. When he looked at Revelo, he saw a company positioned at the intersection of the two things that matter most: world-class engineering talent and the infrastructure that makes AI code generation better.
In Mat's Words
"The world changed fast. When a developer can produce millions of lines of code every day, their value doesn't go down; it goes through the roof. Every lab I've worked with is pouring resources into code generation because that's where the commercial opportunity is. But better models require better feedback from real engineers. That's what Revelo has."
What's Next
Mat will be helping Revelo deepen relationships with leading AI labs and shape our code-gen data infrastructure roadmap.
The companies that figure out the flywheel first will have a massive edge: the best AI coding tools paired with engineers who push both the tools and the technology forward. That's who we're built to serve.
Welcome, Mat.
For press inquiries, contact Steve Patrizi at steve@revelo.com.
