How to Update and Modify Your Custom App After Creation

How to Update Your App Step by Step

Updating a custom app uses the same AI builder interface you used to create it. The process is straightforward and takes just a few minutes for most changes.

Common Modifications You Can Make

The AI can modify any part of your custom app. Here are the most common types of changes people make after the initial build.

Adding New Fields

The most frequent modification is adding new data fields. You might realize after using the app that you need to track additional information, like a status field, a date field, a notes field, or a category selector. Tell the AI what field to add, what type it should be (text, dropdown, date, checkbox, etc.), and where it should appear on the admin page. The AI updates the database structure, the admin page configuration, and any related logic.

Changing the Admin Layout

You can rearrange how fields appear on the edit page, change which columns show in the list view, add or remove filters, enable sorting, or restructure sections. The AI modifies the slug configuration that defines the admin interface without touching the backend logic, so your data and API endpoints are unaffected.

Adding New Commands

If your app needs a new API endpoint, describe what the command should do, what data it should accept, and what it should return. The AI writes the new function, registers it in the app configuration, and ensures it follows the same security and authentication patterns as existing commands. You can start calling the new endpoint immediately after the update.

Modifying Background Job Logic

Background jobs can be added, removed, or changed. You can modify when a job runs, what data it processes, what actions it takes, or how it handles errors. The AI updates both the job function and the schedule configuration. If you need to change a daily job to an hourly job, or add a condition that makes the job skip certain records, just describe the change in the conversation.

Adding AI Features

You can add AI-powered features to an existing app at any time. For example, you might start with a simple data management app and later decide to add automatic categorization using GPT, smart search using embeddings, or AI-generated summaries of records. The AI integrates these features into your existing code and connects them to the appropriate data fields.

Adding Email or SMS Notifications

If your app should send notifications that were not part of the original design, the AI can add email or SMS sending to any command, background job, or webhook handler. Describe when the notification should go out, who should receive it, and what the message should contain.

How the AI Preserves Existing Data and Features

When the AI modifies your app, it works carefully to preserve everything that already exists. Your data stays intact, existing features continue to work, and API endpoints maintain their behavior unless you specifically ask for changes.

The AI achieves this by loading the complete current version of your app before making any changes. It reads the existing PHP functions, the slug configurations, and the app configuration, then makes targeted modifications to only the parts that need to change. Adding a new field does not affect existing fields. Adding a new command does not change how existing commands work. Changing the admin layout does not alter the backend logic.

For database changes specifically, the platform uses a flexible schema approach. Records are stored as JSON objects in DynamoDB, so adding a new field does not require a migration or schema change. Existing records simply do not have the new field yet, and the app handles this gracefully by using default values. When you edit an existing record after adding a field, the new field gets saved along with all the existing data.

Tips for Describing Changes Effectively

The AI understands natural language, so you do not need to use technical terms. But being specific about what you want helps the AI make the right changes on the first try.

• Name the feature or field specifically. Instead of "add some more information," say "add a field called Delivery Date that uses a date picker."
• Describe the behavior you want. Instead of "make it smarter," say "when a new order is created, automatically send an email to the customer with the order details."
• Mention where things should appear. If you want a field on the edit page, say so. If you want a column in the list view, mention that too. The AI can usually guess, but being explicit avoids a follow-up round.
• Reference existing features when relevant. If you want a new field to work like an existing one, say "add a Status field with the same dropdown style as the Priority field."
• Describe the problem if you are not sure of the solution. Instead of prescribing a specific implementation, you can say "customers keep calling to ask about their order status, I need a way to let them check it themselves." The AI will suggest an appropriate feature.

For more guidance on planning and describing apps, see How to Plan Your Custom App Before Building. To understand what the AI generates under the hood, see How the AI Builds and Writes Your App Code.