AWS throws its weight behind Superblocks, signaling a second wave of enterprise vibe coding

Superblocks, a startup focused on vibe coding for business users, announced a multiyear joint marketing agreement with AWS that lets its tool be embedded within AWS customers' private clouds. The deal is a notable validation for the young company—and a clearer signal of where the AI development market is heading.

For enterprise customers, the pitch is straightforward: Apps built with Superblocks will stay inside their AWS environment, with data never leaving the private cloud. Instead of creating external Supabase databases or calling third-party model providers directly, the apps spin up Amazon Aurora databases and integrate with Amazon Bedrock, AWS's AI gateway. That means the apps fall under IT's management, security, and auditing—a direct answer to the shadow-IT problem that has plagued AI adoption.

"We're going to bring it to your data inside your private cloud," Superblocks co-founder and CEO Brad Menezes told TechCrunch. "The big thing about that is data never leaves. It's their AWS account and basically secure with all of the auditing, all of the encryption, all of the network controls."

For AWS, the deal is a way to extend its reach into vibe coding for business users. AWS doesn't yet have its own product aimed at that segment—Kiro is for developers, Quick is more of a copilot—but it can now offer a curated solution through Superblocks. AWS will also help sell the product through its Marketplace, a common pattern for partners.

More than a startup lifeline

For Superblocks, which has 50 employees and raised $60 million in Series A funding from Spark Capital, Kleiner Perkins, Meritech, and Greenoaks, the deal is a major boost. But it's also emblematic of a broader shift: Hyperscalers are pushing enterprises to decouple AI models from the rest of the AI stack—orchestration, security, and app-level tooling—and run that scaffolding on their own clouds.

Microsoft CEO Satya Nadella has been making this case publicly, warning that AI labs may not be trustworthy partners for agent orchestration. Enterprises, meanwhile, are already adopting multiple models, including open-weight options. OpenAI and Anthropic are no longer the default choices for every workload. Vercel reported that open models accounted for 29% of traffic through its AI gateway last month.

"Having a multi-model strategy across big frontier labs, OpenAI, Anthropic, and open source—and I'd say Chinese open source right now, but also U.S. open source is now starting to come up—it's a must-have for the CIO," Menezes said. He added that enterprises are moving away from single-model bets, predicting that "any enterprise that is betting on a single model provider, that executive will be fired."

If that's true, the hyperscalers' strategy makes sense: They can't win the model race outright, but they can own the harness—the layer that enterprises rely on to manage, secure, and integrate multiple models. AWS's partnership with Superblocks is a step toward owning that layer for business users who want to vibe-code, not just developers who use dedicated coding agents.

For developers, the lesson is clear: The vibe-coding trend is moving from experimentation to enterprise governance. Tools like Superblocks, when embedded in a private cloud, offer a way for enterprises to let business users build apps while keeping control—and that could accelerate adoption in regulated industries that have been cautious about sending data to external model providers.