Conversation
Highlights

From Academia to AI Governance: How Breeze ML is Building the Infrastructure for Responsible AI

Most founders chase the obvious opportunities. When ChatGPT exploded onto the scene, countless startups rushed to build the next great language model. But in a recent episode of Category Visionaries, Harry Xu revealed why Breeze ML took the opposite approach, stepping back from the LLM gold rush to tackle a more fundamental challenge: AI governance.

The Contrarian Bet

“As everybody was all in for LLMs, we kind of backed out,” Harry explains. This wasn’t a decision made in isolation – it came from extensive customer discovery. “I just talked to a lot of people. I had tons of conversations with people doing different things in different roles… data scientists… machine learning engineers… VP of engineering… compliance officers… CTOs, CEOs.”

These conversations revealed a critical insight: while everyone was focused on building AI models, nobody was solving the governance problem that would soon affect every AI company. “We clearly saw governance is a bigger problem in years to come,” Harry notes.

The Regulatory Tsunami

The urgency isn’t theoretical. “The EU AI act is already there, and then they’re looking to finalize the law by the end of this year, and then that’s going to come into effect in the year of 2025,” Harry explains. Companies that aren’t prepared face existential risks: “We’re talking about like a huge fine, something like 6% of your annual global revenue, like uncapped.”

Building in Uncharted Territory

The challenge Breeze ML faces is unique: they’re building for a market that exists but hasn’t been defined. “People don’t know what to do yet,” Harry admits. “There are very few tools out there that can help them provide the governance that need.”

Even the experts are still figuring it out. “We talked to a lot of lawyers and privacy attorneys… everybody was talking about auditing AI, auditing models. But in terms of concrete steps, the action items, nobody had a good idea of what to audit.”

This ambiguity creates both challenges and opportunities. While it makes the sales cycle longer, it also means Breeze ML has the chance to define the category. They’re focusing on enterprise sales, building relationships with companies that will soon need these tools to operate legally.

The Go-to-Market Evolution

For founders navigating similar uncertain waters, Harry’s approach offers valuable lessons. Rather than trying to create demand, they’re positioning themselves ahead of inevitable regulatory requirements. Some sectors are already feeling the pressure: “Healthcare is the industry that is facing regulations from FDA… The banks are facing very strict regulations and compliance from SEC.”

This regulatory pressure creates natural entry points. Instead of trying to serve everyone immediately, Breeze ML can focus on sectors where the need is most acute and expand from there.

Building for the Long Game

The vision is clear: “We’ll be the leading platform in AI governance for both the US and EU market,” Harry states. But getting there requires careful timing and preparation.

His key advice for founders? “If I start the signal one more time… I will start paving the road a year before we see the investors, including, for example, assemble the team, clear out the potential IP issues, do the customer survey and then get MVP build.”

This level of preparation becomes especially crucial when building in an emerging category. Without established playbooks to follow, founders need to be even more methodical in their approach.

For B2B tech founders, Breeze ML’s journey offers a masterclass in category creation. While others chase today’s trends, they’re building the infrastructure for tomorrow’s requirements. It’s a reminder that sometimes the biggest opportunities lie not in the technology itself, but in making that technology viable at scale.