The Story of Openlayer: Building the Safety Net for the AI Revolution

Deep within Apple’s most secretive project – the Vision Pro – a team of engineers was growing increasingly frustrated. Not with the groundbreaking technology they were building, but with the tools they had to ensure it worked safely and reliably.

“I was working on the Apple Vision Pro for a long time. I couldn’t even talk about it because it was a secret project,” shares Gabriel Bayomi, CEO of Openlayer, in a recent episode of Category Visionaries. But secrecy wasn’t the challenge that would eventually drive him to start his own company.

The real problem was more fundamental. “Know building models was not the hard part,” Gabriel explains. “The hard part was everything around it. How do you test to make sure it’s good and safe? How do you monitor it in production to make sure the performance is as good as you expect?”

This frustration resonated with his colleagues at Apple. When Gabriel approached them with the idea of starting a company to solve this problem “once and for all,” they didn’t hesitate. Together, they left their prestigious positions at Apple to found Openlayer.

The timing couldn’t have been better. The team applied directly to Y Combinator, inspired by another Brazilian founder’s success story. “He’s the Founder of Brex. He started at the same time that I was in grad school. And it was a big inspiration for me because it was like someone that came from the same country as I did, doing something so cool going through Y Combinator,” Gabriel recalls.

At YC, the team received unexpected advice from Paul Graham himself that would shape their approach. During office hours, Graham challenged the conventional wisdom about product-market fit, telling them “I don’t really believe in this idea of product market fit… It’s all about being default alive, not default dead.”

This perspective helped the team focus on what really mattered – building something users genuinely needed. As machine learning engineers themselves, they understood the disconnect between sophisticated model development tools and the primitive infrastructure for testing and monitoring those models.

Their early go-to-market strategy faced a surprising obstacle – their own engineering backgrounds. “As a software engineer, if you went to a good school… sometimes people are a little bit scared to ask for help, to ask for that new intro or to ask for what they really want,” Gabriel admits. The breakthrough came when they realized that “becoming a Founder is a very humbling experience because we learned, like, we can’t do this alone.”

This mindset shift proved crucial for landing early enterprise customers like eBay. Instead of relying on abstract marketing messages about AI safety, they focused on concrete value propositions. “Instead of going to the abstract idea space of like, we make your AI safe, we try to market things more directly. For example, hey, get alerts when your LLM fails,” Gabriel explains.

Looking ahead, Openlayer’s vision extends far beyond just being another AI infrastructure company. They aim to become “the guardrail of the AI revolution.” Gabriel envisions a future where “people think about OpenAI and cohere and all of these companies building these huge models and think like, okay, what’s our Openlayer stack? To be able to mitigate all the issues that are going to come naturally from deploying it.”

Just as software engineers today wouldn’t dream of deploying code without proper testing and monitoring, Gabriel believes AI development will follow the same path. “The way that people think about code and creating tests, unit tests, or CI CD pipelines, they’re going to be thinking the same on AI, but using Openlayer.”

In an era where AI capabilities are rapidly expanding but safety concerns are mounting, Openlayer’s journey from internal frustration to essential infrastructure mirrors the broader evolution of the AI industry itself – from exciting possibility to practical reality, with safety and reliability at its core.