The promise of Vibecoding was that you could leverage systems like Lovable and Replit AI to launch full-fledged applications from just an idea. But it turns out that writing code is only the first step in the process. And vibe coders already face the standard headaches of maintaining and updating software products.
Fortunately for them, new startups are popping up to fill the gap. On Wednesday, platform engineering startup Shuttle announced $6 million in seed funding to address infrastructure issues that begin once products like Lovable and Cursor end. Investors include former GitHub CEO Thomas Dohmke and Segment founder Calvin French-Owen.
Shuttle takes the code generated by the Vibecoding system, evaluates the best way to deploy it, and presents users with sensible infrastructure packages along with pricing. Once the user agrees, Shuttle can arrange payment and deploy the software directly to the cloud provider with minimal hassle.
It’s been a long journey for Shuttle, which started as part of the Y Combinator class in 2020. Since then, Shuttle has become one of the most popular systems for deploying Rust apps, attracting 20,000 developers across 120,000 deployments with its fast, zero-configuration approach. With this new round of funding, the company plans to expand its funding into all programming languages and AI coding systems.
As CEO and co-founder Nodar Daneliya explains, agent AI systems have made it much easier to cross the barriers between different programming systems. This means you can deploy systems like Shuttle to all your programming systems at once. “AI is erasing the boundaries between different language ecosystems,” Daneliya says. “For us, it’s the perfect time[to scale up]because we’ve been in this back-end development space for years.”
In practice, this means building an agent interface for platform management. This will allow users to provision databases and purchase cloud hosting using the same natural language prompts they used to vibecode their apps. On the backend, it also means building interconnections with cloud providers and coding systems so that agents can get all the context they need.
“Essentially, we created this specification that acts as a middle layer between what humans can review and what AI can understand,” Daneliya told TechCrunch. “Spec-driven development is becoming the go-to way of doing things, and there’s no reason why it can’t apply to infrastructure as well.”
