Ray AI

Ray AI is a SaaS management tool for vibe coders, allowing users to validate SaaS ideas, manage their app flows, issues, features, and public roadmaps with ease, shipping high-quality SaaS apps.

Comment

Inspiration

I was excited to join the Bolt hackathon, so many ideas to build, and enough time as well. It was an opportunity to change my life for the better. But as a first-time founder-to-be, things get complicated quickly when building SaaS, especially with AI tools such as Bolt. You lose track of what's important, you forget the main goal of the app, the flow gets out of hand, hard to manage the issues, iterate, and launch a high quality vibe-coded app.

I lost all of my credits before I could finish the project, and gave up on my dream. With the weekend of Bolt for free, I got enough motivation to try again, and came up with a more solid plan for my app.

What it does

Initially, I was building a SaaS validation app, but realizing how it was hard to maintain the app flow, issues I am facing, gathering feedback from users, I decided to develop a SaaS funnel.

Ray AI allows you to add your SaaS ideas, and our AI validates everything behind the scenes with AI agents. When you are happy with your validation, you can proceed with building the app. Ray allows you to add projects with fixed types, such as mobile, backend, front-end, and also add your tech stack. The main feature of Ray AI is project flow management with AI, allowing you to generate app flows, PRDs from app flows, documentatio,n and implementation prompts for AI, which allows you to vibe code consistently high-quality apps.

I also integrated issue tracking, public road map for users to see what you are working on, and a lot of tiny easter-egg features.

How we built it

Over 90% of Ray AI was built with Bolt.new, using over 40 Million credits provided during the hackathon. I connected github to fix the BetterAuth integration and attempting subscription with Polar.

I was amazed how well Bolt handled Convex.dev as a database, everything worked perfectly as expected, just came across few bugs here and there while running the app, but nothing I couldn't fix.

Challenges we ran into

The main challenge was Bolt was not being able to handle the project as it gets bigger, so many unused components and code. I had to do a bit of cleanup to fix that. Connecting every other feature including AI was very easy with Bolt.

Accomplishments that we're proud of

I feel proud to have been able to build a complete SaaS MVP, and hoping this will grow into a business I can be proud of. The inspiration I get from other indie hackers and YouTubers is very amazing and even though I lost hope when I ran out of credits, I am grateful we got this far.

What we learned

AI tools are here to stay, and the sooner you learn to use them effeciently, the better. I renewed Bolt subscription because I saw how useful it can be in my daily life.

What's next for Ray AI

I am hoping to add waitlist, in-app feedback SDKs, connecting with Github and Slack to sync issues and notifications. I am also deciding on the subscription structure to go for, but overall, this is a product destined to grow.

Built With

  • ai-sdk
  • betterauth
  • convex
  • nextjs
  • polar
  • tailwind
  • typescript
Share this project:

Updates