Ramp and Stripe built internal vibe-coding tools instead of buying them
Two well-known fintech companies quietly built their own AI-powered internal app builders rather than paying for Lovable, Replit, or Retool. Now one of them has shipped the tool as an external product.
Ramp’s Glass
Ramp built Glass, an internal vibe-coding platform that lets non-engineers create and share complex applications connected to internal data sources. A four-person team built it. The tool has spread far enough inside Ramp that onboarding to Glass and building a simple app is now part of new hire onboarding for non-engineers.
One example: a marketing employee built a custom CRM with 50-plus fields and filters entirely inside Glass. Ramp has since launched Glass as an external product, describing it as “An AI workspace where knowledge compounds.”
Stripe’s Harbor
Stripe published details on its own equivalent, called Harbor, which Stripe describes as an “AI-assisted prototyping tool.” A single engineer built Harbor v1 in roughly six weeks.
Why build instead of buy?
Gergely Orosz, writing in The Pragmatic Engineer, outlines six reasons companies are choosing to build their own:
- Low investment. Stripe used one engineer for six weeks. Ramp used two engineers. For a company with hundreds of engineers, that is a manageable spend.
- Internal data stays internal. Routing proprietary data through a third-party tool like Lovable or Replit carries real risk. A self-hosted tool removes that exposure.
- Design system consistency. Both Ramp and Stripe specifically engineered their tools to follow internal design guidelines. Off-the-shelf tools do not do this by default.
- Broader adoption. Internal tools built for the actual internal context tend to get used more than vendor tools that require customization and compliance reviews.
- No vendor dependency. The effort to negotiate, review, and onboard a vendor is comparable to just building a focused internal solution.
- Engineer upskilling. Building and maintaining AI infrastructure is hard to hire for right now. Building these tools in-house trains existing engineers on AI infra.
The operator takeaway
You don’t need 500 engineers to take this lesson. If you’re running a small team with internal data you can’t expose to SaaS tools, a focused internal AI app builder built on top of an existing LLM API is worth the exploration. The build cost at Stripe was one engineer for six weeks. The payoff was company-wide adoption and a tool tailored to their exact stack.
I’m founder of BizStack at Cagri Sarigoz LLC and a passionate advocate for entrepreneurs.
With over 14 years in tech, marketing, and AI, including my role as Head of SEO at CitizenShipper and co-founder of TaleBot at Intale AI, I’m dedicated to sharing genuine, useful product insights and tips.
At BizStack, I aim to cut through the digital noise to provide clear, actionable advice.
And more than all else, I’m a father to a (always) little girl and a husband.
Contact me at [email protected] for assistance.
