Post image

“Neon’s speed of provisioning and serverless scale-to-zero is critical for us. We can serve users iterating on quick ideas efficiently while also supporting them as they scale, without making them think about database setup.” (Dhruv Amin, Co-founder at Create.xyz)

These are wild times. It’s now possible to describe a fancy app in a single text prompt, e.g.,

“Make an AI recipe generator with ChatGPT that lets me enter a food, generate recipes and save my favorites to a gallery for later”

And watch a complete app materialize in seconds:

If you’ve been playing with Replit, Bolt or Lovable, Create is next on your list.  

From Text to Full Stack, Fast 

Create makes building production-ready apps accesible to everyone. There’s two core elements that make users love it:

  • End-to-end
  • Speed

Speed

Rather than make users wait, Create’s agent knows how to make targeted updates to the codebase. This is especially important for people building more sophisticated applications that require many iterations. The AI workflow that produces truly useful apps is simple: see your app, test it, and refine it on the fly—but this must happen without delays.

End-to-end   

When you build with agents, the apps are often almost usable UI prototypes but not quite real—they’re missing something essential and it’s hard to make them go end to end. Create aims to change this. When you give Create an instruction, it turns it into the right pages, components, backend functions, and database setup. You see them appear visually in real-time.    

Create can also automatically add 50+ built-in integrations, and makes it easy to add your own integrations with its backend functions. The entire workflow of adding and managing integrations is abstracted away—just describe what you need, and the agent handles the setup. Here’s a shortlist of what Create can add to your app:

  • AI & LLMs. Use OpenAI, Anthropic, Gemini, Llama, and other foundational models to add AI features like content generation, chatbots, or classification. 
  • Image generation. Call DALL·E, Stable Diffusion, and other models.
  • Payments. Ask Create to add Stripe to start collecting payments for your product right away.

Since Create is generating code for your app under the hood, you have control over the UI down to the pixel and exact logic.  And when you’re ready to go live, Create deploys it on developer-grade infra.

Enter Neon: Powering the Database Behind the Scenes

Among all this goodness, Create bakes-in a Postgres database from the start—and that’s where Neon comes in. Neon is a serverless Postgres database built for fast, iterative workflows, making it a perfect fit for AI agents: 

  • Provisioning happens almost instantly in Neon. AI agents can spin up fully functional Postgres database in less than a second. 
  • Neon autoscales infrastructure, starting from zero. Neon is serverless, with a pay-as-you-go model. It automatically scales compute and storage as the app demands it, which is critical for agents: this pricing model ensures very low costs when databases are idle (for the AI company) and good scalability (for the end-user) as usage grows. 

With Neon, Postgres management becomes just another API call. No manual provisioning, no downtime, no headaches—just instant, scalable Postgres that works behind the scenes.

But Neon has another big advantage for agentic workflows: it supports branches, aka isolated copies of your data and schema that are provisioned instantaneously. Neon’s branching model allows Create to solve a big pain when building AI agents: how to handle schema migrations. 

“When users want to iterate on their app, we can branch off their data with Neon and test schema migrations in isolation. We can even auto-fix issues. Neon’s branching model is a huge win for iteration speed” (Dhruv Amin, Co-founder at Create.xyz)

Without a structured approach, schema migrations can introduce downtime, which hurts the user experience especially when updates are meant to happen dynamically in response to user prompts. Neon’s branching allows Create to apply an iterative workflow for schema migrations: 

  1. User proposes schema changes. When a user prompts Create with “Add payments to the app” or “Let users save a twitter link to their profile,” the system automatically generates the necessary database migration and applies it to an isolated Neon branch, which acts as a separate environment where the changes can be tested first. 
  2. Changes are tested. This separate environment is ready immediately, and it’s a perfect copy of the main database (schema + data). The user and the agent can validate migrations in parallel in this environment, without affecting the original version of the app. 
  3. Changes are applied. Once the migration is verified, the agent can apply the changes back into production.

Start Building

You can sign up to Create’s Free Plan and start exploring the agent immediately. Share what you’ve built by mentioning them on X: https://x.com/create_xyz

If you’re a company building an AI agent, let’s chat. We have a Postgres API for you.